Oct. 2023, Information Processing Society of Japan, IPSJ-CS Outstanding Achievement and Contribution Award　　　　　　　　　　　　　　　

Sep. 2017, 情報処理学会数理モデル化と問題解決研究会, 功績賞　　　　　　　　　　　　　　　
棟朝 雅晴

Mar. 2009, The 10-th WSEAS international conference on Evolutionary Computation, Best paper award　　　　　　　　　　　　　　　
MUNETOMO Masaharu

SFE-EANDS: a simple, normalized distance-based selectionfast, and efficient algorithm with external archive and normalized distance-based selection for high-dimensional feature selection.
Rui Zhong, Yang Cao, Essam H. Houssein, Jun Yu 0012, Masaharu Munetomo
Cluster Computing, 28, 5, 285, 285, Oct. 2025
Scientific journal

Adjacent Distance Matrix-Based Competitive Swarm Optimizer
Yang Cao, Rui Zhong, Jun Yu, Masaharu Munetomo
Lecture Notes in Computer Science, 38, 51, Springer Nature Switzerland, 17 Apr. 2025
In book

Competitive differential evolution with knowledge inheritance for single-objective human-powered aircraft design
Rui Zhong, Yang Cao, Enzhi Zhang, Masaharu Munetomo
The Journal of Supercomputing, 81, 5, 721, 721, Springer Science and Business Media LLC, 10 Apr. 2025
Scientific journal

HHDE: a hyper-heuristic differential evolution with novel boundary repair technique for complex optimization
Rui Zhong, Shilong Zhang, Jun Yu, Masaharu Munetomo
The Journal of Supercomputing, 81, 5, 696, 696, Springer Science and Business Media LLC, 04 Apr. 2025
Scientific journal

Vision Transformer-Based Meta Loss Landscape Exploration with Actor-Critic Method
Enzhi Zhang, Rui Zhong, Xingbang Du, Mohamed Wahib, Masaharu Munetomo
Communications in Computer and Information Science, 291, 305, Springer Nature Switzerland, 26 Mar. 2025
In book

Hyper-heuristic Differential Evolution with Novel Boundary Repair for Numerical Optimization
Rui Zhong, Jun Yu, Masaharu Munetomo
Communications in Computer and Information Science, 264, 277, Springer Nature Switzerland, 26 Mar. 2025
In book

Introducing Competitive Mechanism to Differential Evolution for Numerical Optimization
Rui Zhong, Yang Cao, Enzhi Zhang, Masaharu Munetomo
Communications in Computer and Information Science, 251, 263, Springer Nature Switzerland, 26 Mar. 2025
In book

LLMOA: A novel large language model assisted hyper-heuristic optimization algorithm
Rui Zhong, Abdelazim G. Hussien, Jun Yu, Masaharu Munetomo
Advanced Engineering Informatics, 64, 103042, 103042, Elsevier BV, Mar. 2025
Scientific journal

Forecasting Renewable energy and electricity consumption using evolutionary hyperheuristic algorithm
Yang Cao, Jun Yu, Rui Zhong, Masaharu Munetomo
Scientific Reports, 15, 1, Springer Science and Business Media LLC, 20 Jan. 2025
Scientific journal

Vision transformer-based meta loss landscape exploration with actor-critic method
Enzhi Zhang, Rui Zhong, Xingbang Du, Mohamed Wahib, Masaharu Munetomo
The Journal of Supercomputing, 81, 1, 350, 350, Springer Science and Business Media LLC, 28 Dec. 2024
Scientific journal

Hierarchical Adaptive Differential Evolution with Local Search for Extreme Learning Machine
Rui Zhong, Yang Cao, Jun Yu, Masaharu Munetomo
Lecture Notes in Computer Science, 235, 246, Springer Nature Singapore, 21 Aug. 2024
In book

Large Language Model Assisted Adversarial Robustness Neural Architecture Search
Rui Zhong, Yang Cao, Jun Yu, Masaharu Munetomo
2024 6th International Conference on Data-driven Optimization of Complex Systems (DOCS), 433, 437, IEEE, 16 Aug. 2024
International conference proceedings

Optimization of Electricity Consumption Forecasting Models via Hyper-Heuristic Algorithm
Yang Cao, Rui Zhong, Jun Yu, Masaharu Munetomo
2024 6th International Conference on Data-driven Optimization of Complex Systems (DOCS), 114, 120, IEEE, 16 Aug. 2024
International conference proceedings

GeminiDE: A Novel Parameter Adaptation Scheme in Differential Evolution
Rui Zhong, Shilong Zhang, Jun Yu, Masaharu Munetomo
2024 6th International Conference on Data-driven Optimization of Complex Systems (DOCS), 33, 38, IEEE, 16 Aug. 2024
International conference proceedings

Validation Loss Landscape Exploration with Deep Q-Learning
Enzhi Zhang, Rui Zhong, Masaharu Munetomo, Mohamed Wahib
2024 International Joint Conference on Neural Networks (IJCNN), 1, 9, IEEE, 30 Jun. 2024
International conference proceedings

Meta generative image and text data augmentation optimization.
Enzhi Zhang, Bochen Dong, Mohamed Wahib, Rui Zhong, Masaharu Munetomo
J. Supercomput., 80, 9, 12644, 12662, Jun. 2024
Scientific journal

Evolutionary multi-mode slime mold optimization: a hyper-heuristic algorithm inspired by slime mold foraging behaviors.
Rui Zhong, Enzhi Zhang, Masaharu Munetomo
J. Supercomput., 80, 9, 12186, 12217, Jun. 2024
Scientific journal

Cooperative coevolutionary surrogate ensemble-assisted differential evolution with efficient dual differential grouping for large-scale expensive optimization problems
Rui Zhong, Enzhi Zhang, Masaharu Munetomo
Complex and Intelligent Systems, 10, 2, 2129, 2149, Apr. 2024
Scientific journal

SRIME: a strengthened RIME with Latin hypercube sampling and embedded distance-based selection for engineering optimization problems.
Rui Zhong, Jun Yu, Chao Zhang 0030, Masaharu Munetomo
Neural Comput. Appl., 36, 12, 6721, 6740, Apr. 2024
Scientific journal

Adaptive Patching for High-resolution Image Segmentation with Transformers.
Enzhi Zhang, Isaac Lyngaas, Peng Chen 0035, Xiao Wang 0004, Jun Igarashi, Yuankai Huo, Masaharu Munetomo, Mohamed Wahib
SC, 76, 76, 2024
International conference proceedings

Cooperative coevolutionary differential evolution with adjacent intensity matrix with linkage identification for large-scale optimization problems in noisy environments
Rui Zhong, Binnan Tu, Enzhi Zhang, Masaharu Munetomo
Evolutionary Intelligence, 2024
Scientific journal

Learning from the Past Training Trajectories: Regularization by Validation.
Enzhi Zhang, Mohamed Wahib, Rui Zhong, Masaharu Munetomo
J. Adv. Comput. Intell. Intell. Informatics, 28, 1, 67, 78, Jan. 2024
Scientific journal

Q-learning based vegetation evolution for numerical optimization and wireless sensor network coverage optimization
Rui Zhong, Fei Peng, Jun Yu, Masaharu Munetomo
Alexandria Engineering Journal, 87, 148, 163, Jan. 2024
Scientific journal

Surrogate Ensemble-Assisted Hyper-Heuristic Algorithm for Expensive Optimization Problems
Rui Zhong, Jun Yu, Chao Zhang, Masaharu Munetomo
International Journal of Computational Intelligence Systems, 16, 1, Dec. 2023
Scientific journal

An evolutionary robotics approach to a multi-legged robotic swarm in a rough terrain environment
Daichi Morimoto, Haruhi Tsukamoto, Motoaki Hiraga, Kazuhiro Ohkura, Masaharu Munetomo
Artificial Life and Robotics, 28, 4, 661, 668, Nov. 2023
Scientific journal

Vegetation Evolution with Dynamic Maturity Strategy and Diverse Mutation Strategy for Solving Optimization Problems
Rui Zhong, Fei Peng, Enzhi Zhang, Jun Yu, Masaharu Munetomo
Biomimetics, 8, 6, Oct. 2023
Scientific journal

Training Knowledge Inheritance Through Deep Q-Net
Enzhi Zhang, Ruqin Wang, Mohamed Wahib, Rui Zhong, Masaharu Munetomo
Conference Proceedings - IEEE International Conference on Systems, Man and Cybernetics, 899, 904, 2023
International conference proceedings

Meta Generative Data Augmentation Optimization
Enzhi Zhang, Bochen Dong, Mohamed Wahib, Rui Zhong, Masaharu Munetomo
Proceedings - 2023 Congress in Computer Science, Computer Engineering, and Applied Computing, CSCE 2023, 2218, 2225, 2023
International conference proceedings

Evolutionary Multi-Mode Slime Mould Optimization: A Hyper-Heuristic Algorithm Inspired by Slime Mould Foraging Behaviors
Rui Zhong, Enzhi Zhang, Masaharu Munetomo
Proceedings - 2023 Congress in Computer Science, Computer Engineering, and Applied Computing, CSCE 2023, 2153, 2160, 2023
International conference proceedings

Adjacent Intensity Matrix with Linkage Identification for Large-Scale Optimization in Noisy Environments.
Rui Zhong, Binan Tu, Enzhi Zhang, Masaharu Munetomo
CEC, 1, 10, 2023
International conference proceedings

A Hierarchical Cooperative Coevolutionary Approach to Solve Very Large-Scale Traveling Salesman Problem.
Rui Zhong, Enzhi Zhang, Masaharu Munetomo
OLA, 74, 84, 2023
International conference proceedings

Cooperative coevolutionary differential evolution with linkage measurement minimization for large-scale optimization problems in noisy environments
Rui Zhong, Enzhi Zhang, Masaharu Munetomo
COMPLEX & INTELLIGENT SYSTEMS, Jan. 2023
English, Scientific journal

Cooperative Coevolutionary NSGA-II with Linkage Measurement Minimization for Large-Scale Multi-objective Optimization.
Rui Zhong, Masaharu Munetomo
EMO, 43, 55, 2023
International conference proceedings

Generating collective behavior of a multi-legged robotic swarm using an evolutionary robotics approach
Daichi Morimoto, Motoaki Hiraga, Naoya Shiozaki, Kazuhiro Ohkura, Masaharu Munetomo
Artificial Life and Robotics, 27, 4, 751, 760, Nov. 2022
Scientific journal

Generating collective behavior of a multi-legged robotic swarm using an evolutionary robotics approach
Daichi Morimoto, Motoaki Hiraga, Naoya Shiozaki, Kazuhiro Ohkura, Masaharu Munetomo
ARTIFICIAL LIFE AND ROBOTICS, 27, 4, 751, 760, Nov. 2022
English, Scientific journal

Report on Open Space Discussion 2021
能島, 裕介, 高木, 英行, 棟朝, 雅晴, 濱田, 直希, 西原, 慧, 高玉, 圭樹, 佐藤, 寛之, 桐淵, 大貴, 宮川, みなみ
進化計算学会論文誌, 13, 1, 1, 9, 進化計算学会, 13 Jul. 2022
Japanese, Scientific journal, This paper is a report on Open Space Discussion (OSD) held in Evolutionary Computation Symposium 2021. The purpose of OSD is to share and discuss problems at hand and future research targets related to evolutionary computation. Discussion topics are voluntarily proposed by some of the participants, and other participants freely choose one to join in the discussion. Through free discussions based on the open space technology framework, it is expected that participants will have new research ideas and start some collaborations. This paper gives the concept of OSD and introduces six topics discussed this year. This paper also shows the responses to the questionnaire on OSD for future discussions, collaborations, and related events.

Generating and Analyzing Collective Step-Climbing Behavior in a Multi-legged Robotic Swarm
Daichi Morimoto, Motoaki Hiraga, Kazuhiro Ohkura, Masaharu Munetomo
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 13491 LNCS, 324, 331, 2022
International conference proceedings

Learning from the Past: Regularization by Validation.
Enzhi Zhang, Mohamed Wahib, Masaharu Munetomo
SCIS/ISIS, 1, 8, 2022
International conference proceedings

Many-Constraint and Many-Objective optimization with Bias Index for Intercloud Multi-Workflow Resource Provisioning.
Courtney Powell, Katsunori Miura, Masaharu Munetomo
SCIS/ISIS, 1, 8, IEEE, 29 Nov. 2022
International conference proceedings

Accelerating the Evolutionary Algorithms by Gaussian Process Regression with ε-greedy acquisition function.
Rui Zhong, Enzhi Zhang, Masaharu Munetomo
CoRR, abs/2210.06814, 2022
Scientific journal

Accelerating the Genetic Algorithm for Large-scale Traveling Salesman Problems by Cooperative Coevolutionary Pointer Network with Reinforcement Learning.
Rui Zhong, Enzhi Zhang, Masaharu Munetomo
CoRR, abs/2209.13077, 2022
Scientific journal

Cooperative coevolutionary Modified Differential Evolution with Distance-based Selection for Large-Scale Optimization Problems in noisy environments through an automatic Random Grouping.
Rui Zhong, Masaharu Munetomo
CoRR, abs/2209.00777, 2022
Scientific journal

Accelerating differential evolution algorithm with Gaussian sampling based on estimating the convergence points.
Rui Zhong, Masaharu Munetomo
CoRR, abs/2208.14619, 2022
Scientific journal

Cooperative coevolutionary hybrid NSGA-II with Linkage Measurement Minimization for Large-scale Multi-objective optimization.
Rui Zhong, Masaharu Munetomo
CoRR, abs/2208.13415, 2022
Scientific journal

Optimal answer generation by equivalent transformation incorporating multi-objective genetic algorithm.
Katsunori Miura, Courtney Powell, Masaharu Munetomo
Soft Computing, 26, 19, 10535, 10546, 2022
Scientific journal

Evolving collective step-climbing behavior in multi-legged robotic swarm.
Daichi Morimoto, Motoaki Hiraga, Naoya Shiozaki, Kazuhiro Ohkura, Masaharu Munetomo
Artif. Life Robotics, 27, 2, 333, 340, 2022
Scientific journal

Distribution system for japanese synthetic population data with protection level
Tadahiko Murata, Susumu Date, Yusuke Goto, Toshihiro Hanawa, Takuya Harada, Manabu Ichikawa, Hao Lee, Masaharu Munetomo, Akiyoshi Sugiki
Proceedings - International Conference on Machine Learning and Cybernetics, 2020-December, 187, 193, 02 Dec. 2020
International conference proceedings

GTOPX Space Mission Benchmarks.
Martin Schlueter, Mehdi Neshat, Mohamed Wahib, Masaharu Munetomo, Markus Wagner 0007
CoRR, abs/2010.07517, 2020
Scientific journal

Multi-objective global optimization for interplanetary space trajectory design
Martin Schlueter, Masaharu Munetomo
AIP Conference Proceedings, 2070, American Institute of Physics Inc., 12 Feb. 2019
English, International conference proceedings

Distribution of Synthetic Populations of Japan for Social Scientists and Social Simulation Researchers.
Tadahiko Murata, Takuya Harada, Manabu Ichika Wa, Yusuke Goto, Lee Hao, Susumu Date, Masaharu Munetomo, Akiyoshi Sugiki
1, 5, 2019, [Peer-reviewed]
International conference proceedings

Constrained Multi-objective Optimization Method for Practical Scientific Workflow Resource Selection.
Courtney Powell, Katsunori Miura, Masaharu Munetomo
Evolutionary Multi-Criterion Optimization - 10th International Conference, EMO 2019, East Lansing, MI, USA, March 10-13, 2019, Proceedings, 683, 694, Springer, 2019, [Peer-reviewed]

Network Structural Vulnerability: A Multiobjective Attacker Perspective.
Luca Faramondi, Gabriele Oliva, Stefano Panzieri, Federica Pascucci, Martin Schlueter, Masaharu Munetomo, Roberto Setola
IEEE Trans. Syst. Man Cybern. Syst., 49, 10, 2036, 2049, 2019, [Peer-reviewed]
Scientific journal

A Mixed-Integer Extension for ESA's Cassini1 Space Mission Benchmark.
Martin Schlueter, Masaharu Munetomo
IEEE Congress on Evolutionary Computation, CEC 2019, Wellington, New Zealand, June 10-13, 2019, 912, 919, IEEE, 2019, [Peer-reviewed]
International conference proceedings

Massively parallelized co-evaluation for many-objective space trajectory optimization.
Martin Schlueter, Masaharu Munetomo
Proceedings of the Genetic and Evolutionary Computation Conference Companion, GECCO 2018, Kyoto, Japan, July 15-19, 2018, 306, 307, ACM, 2018, [Peer-reviewed]

Towards a small diverse pareto-optimal solutions set generator for multiobjective optimization problems.
Courtney Powell, Katsunori Miura, Masaharu Munetomo
Proceedings of the Genetic and Evolutionary Computation Conference Companion, GECCO 2018, Kyoto, Japan, July 15-19, 2018, 298, 299, ACM, 2018, [Peer-reviewed]

Introducing a linkage identification considering non-monotonicity to multi-objective evolutionary optimization with decomposition for real-valued functions.
Kousuke Izumiya, Masaharu Munetomo
Proceedings of the Genetic and Evolutionary Computation Conference Companion, GECCO 2018, Kyoto, Japan, July 15-19, 2018, 179, 180, ACM, 2018, [Peer-reviewed]

A surrogate-assisted selection scheme for genetic algorithms employing multi-layer neural networks.
Masaki Fujiwara, Masaharu Munetomo
Proceedings of the Genetic and Evolutionary Computation Conference Companion, GECCO 2018, Kyoto, Japan, July 15-19, 2018, 41, 42, ACM, 2018, [Peer-reviewed]

Optimal Cloud Resource Selection Method Considering Hard and Soft Constraints and Multiple Conflicting Objectives.
Courtney Powell, Katsunori Miura, Masaharu Munetomo
11th IEEE International Conference on Cloud Computing, CLOUD 2018, San Francisco, CA, USA, July 2-7, 2018, 831, 835, IEEE Computer Society, 2018, [Peer-reviewed]

Optimal and Feasible Cloud Resource Configurations Generation Method for Genomic Analytics Applications.
Katsunori Miura, Courtney Powell, Masaharu Munetomo
2018 IEEE International Conference on Cloud Computing Technology and Science, CloudCom 2018, Nicosia, Cyprus, December 10-13, 2018, 137, 144, IEEE Computer Society, 2018, [Peer-reviewed]

A Level-Wise Load Balanced Scientific Workflow Execution Optimization Using NSGA-II
Phyo Thandar Thant, Courtney Powell, Martin Schlueter, Masaharu Munetomo
Proceedings - 2017 17th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing, CCGRID 2017, 882, 889, Institute of Electrical and Electronics Engineers Inc., 10 Jul. 2017, [Peer-reviewed]
English, International conference proceedings

Multi-objective Evolutionary optimization based on Decomposition with Linkage Identification considering monotonicity
Kousuke Izumiya, Masaharu Munetomo
2017 IEEE Congress on Evolutionary Computation, CEC 2017 - Proceedings, 905, 912, Institute of Electrical and Electronics Engineers Inc., 05 Jul. 2017, [Peer-reviewed]
English, International conference proceedings

MIDACO Parallelization Scalability on 200 MINLP Benchmarks
Martin Schlueter, Masaharu Munetomo
Journal of Artificial Intelligence and Soft Computing Research, 7, 3, 171, 181, De Gruyter Open Ltd, 01 Jul. 2017, [Peer-reviewed]
English, Scientific journal

Multiobjective Level-Wise Scientific Workflow Optimization in IaaS Public Cloud Environment
Phyo Thandar Thant, Courtney Powell, Martin Schlueter, Masaharu Munetomo
SCIENTIFIC PROGRAMMING, 2017, [Peer-reviewed]
English, Scientific journal

Numerical Optimization of ESA's Messenger Space Mission Benchmark
Martin Schlueter, Mohamed Wahib, Masaharu Munetomo
APPLICATIONS OF EVOLUTIONARY COMPUTATION, EVOAPPLICATIONS 2017, PT I, 10199, 725, 737, 2017, [Peer-reviewed]
English, International conference proceedings

Multi-Objective Hadoop Configuration Optimization using Steady-State NSGA-II
Phyo Thandar Thant, Courtney Powell, Akiyoshi Sugiki, Masaharu Munetomo
2016 JOINT 8TH INTERNATIONAL CONFERENCE ON SOFT COMPUTING AND INTELLIGENT SYSTEMS (SCIS) AND 17TH INTERNATIONAL SYMPOSIUM ON ADVANCED INTELLIGENT SYSTEMS (ISIS), 293, 298, 2016, [Peer-reviewed]
English, International conference proceedings

Evaluation of Three Steady-State NSGA-III Offspring Selection Schemes for Many-Objective Optimization
Courtney Powell, Masaharu Munetomo, Phyo Thandar Thant
2016 JOINT 8TH INTERNATIONAL CONFERENCE ON SOFT COMPUTING AND INTELLIGENT SYSTEMS (SCIS) AND 17TH INTERNATIONAL SYMPOSIUM ON ADVANCED INTELLIGENT SYSTEMS (ISIS), 166, 171, 2016, [Peer-reviewed]
English, International conference proceedings

Intercloud Brokerages based on PLS Method for deploying Infrastructures for Big Data Analytics
Katsunori Miura, Tazro Ohta, Courtney Powell, Masaharu Munetomo
2016 IEEE INTERNATIONAL CONFERENCE ON BIG DATA (BIG DATA), 2097, 2102, 2016, [Peer-reviewed]
English, International conference proceedings

A Predicate Logic-defined Specification Method for Systems Deployed by Intercloud Brokerages
Katsunori Miura, Masaharu Munetomo
2016 IEEE INTERNATIONAL CONFERENCE ON CLOUD ENGINEERING WORKSHOP (IC2EW), 172, 177, 2016, [Peer-reviewed]
English, International conference proceedings

Development of a Multi-player Interactive Genetic Algorithm-based 3D Modeling System for Glasses
Takahito Seyama, Masaharu Munetomo
2016 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC), 846, 852, 2016, [Peer-reviewed]
English, International conference proceedings

Numerical Assessment of the Parallelization Scalability on 200 MINLP Benchmarks
Martin Schlueter, Masaharu Munetomo
2016 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC), 830, 837, 2016, [Peer-reviewed]
English, International conference proceedings

Screening for FtsZ Dimerization Inhibitors Using Fluorescence Cross-Correlation Spectroscopy and Surface Resonance Plasmon Analysis
Shintaro Mikuni, Kota Kodama, Akira Sasaki, Naoki Kohira, Hideki Maki, Masaharu Munetomo, Katsumi Maenaka, Masataka Kinjo
PLOS ONE, 10, 7, e0130933, Jul. 2015, [Peer-reviewed]
English, Scientific journal

Distributed Denial of Services Attack Protection System with Genetic Algorithms on Hadoop Cluster Computing Framework
Masataka Mizukoshi, Masaharu Munetomo
2015 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC), 1575, 1580, 2015, [Peer-reviewed]
English, International conference proceedings

アカデミッククラウド実現にむけたクラウド支援サービス (インターネットアーキテクチャ)
合田 憲人, 山地 一禎, 中村 素典, 横山 重俊, 吉岡 信和, 政谷 好伸, 西村 浩二, 棟朝 雅晴
電子情報通信学会技術研究報告 = IEICE technical report : 信学技報, 114, 236, 1, 5, 一般社団法人電子情報通信学会, 07 Oct. 2014
Japanese

Human-based Genetic Algorithm for Facilitating Practical Use of Data in the Internet
Ryosuke Hasebe, Rina Kouda, Kei Ohnishi, Masaharu Munetomo
2014 JOINT 7TH INTERNATIONAL CONFERENCE ON SOFT COMPUTING AND INTELLIGENT SYSTEMS (SCIS) AND 15TH INTERNATIONAL SYMPOSIUM ON ADVANCED INTELLIGENT SYSTEMS (ISIS), 1327, 1332, 2014
English, International conference proceedings

Design of an SSO Authentication Infrastructure for Heterogeneous Inter-cloud Environments
Courtney Powell, Takashi Aizawa, Masaharu Munetomo
2014 IEEE 3RD INTERNATIONAL CONFERENCE ON CLOUD NETWORKING (CLOUDNET), 102, 107, 2014, [Peer-reviewed]
English, International conference proceedings

Parallelization for Space Trajectory Optimization
Martin Schlueter, Masaharu Munetomo
2014 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC), 832, 839, 2014, [Peer-reviewed]
English, International conference proceedings

An adaptive parameter binary-real coded genetic algorithm for constraint optimization problems: Performance analysis and estimation of optimal control parameters
Omar Arif Abdul-Rahman, Masaharu Munetomo, Kiyoshi Akama
INFORMATION SCIENCES, 233, 54, 86, Jun. 2013, [Peer-reviewed]
English, Scientific journal

Parallelization strategies for evolutionary algorithms for MINLP
Martin Schlueter, Masaharu Munetomo
2013 IEEE Congress on Evolutionary Computation, CEC 2013, 635, 641, IEEE, 2013, [Peer-reviewed]
English, International conference proceedings

Towards thought control of next-generation wearable computing devices
Courtney Powell, Masaharu Munetomo, Martin Schlueter, Masataka Mizukoshi
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 8211, 427, 438, Springer, 2013, [Peer-reviewed]
English, International conference proceedings

A scalable infrastructure of interactive evolutionary computation to evolve services online with data
Masaharu Munetomo, Shintaro Bando
Proceedings - 2013 IEEE International Conference on Big Data, Big Data 2013, 28, IEEE, 2013, [Peer-reviewed]
English, International conference proceedings

arGA: Adaptive resolution micro-genetic algorithm with tabu search to solve MINLP problems using GPU
Asim Munawar, Mohamed Wahib, Masaharu Munetomo, Kiyoshi Akama
Natural Computing Series, 46, 83, 104, Springer Verlag, 2013, [Peer-reviewed]
English, In book

Constructing a robust services-oriented inter-cloud portal based on an autonomic model and FOSS
Courtney Powell, Masaharu Munetomo, Attia Wahib, Takashi Aizawa
Proceedings - 2013 IEEE/ACM 6th International Conference on Utility and Cloud Computing, UCC 2013, 458, 463, IEEE Computer Society, 2013, [Peer-reviewed]
English, International conference proceedings

Design of Authentication System for High Performance Distributed Computing Environment
Kento Aida, Manabu Higashida, Eisaku Sakane, Hirofumi Amano, Katsushi Kobayashi, Masaharu Munetomo, Ryusuke Egawa, Osamu Tatebe, Yoshikazu Kamoshida, Shin'ichiro Takizawa, Toru Nagai, Takeshi Iwashita, Yutaka Ishikawa
先進的計算基盤システムシンポジウム論文集, 5, 2012, 227, 236, Information Processing Society of Japan (IPSJ), 09 May 2012
Japanese, This paper presents design of the authentication system for the High Performance Computing Infrastructure (HPCI), which is currently deployed by the Ministry of Education, Culture, Sports, Science and Technology. The presented authentication system enables single sign-on to computers and shared storages on HPCI by utilizing the authentication mechanism on the Grid, "Grid Security Infrastructure (GSI)", and the identity federation mechanism, "Shibboleth". This paper also presents the experiments conducted on the testbed for the presented authentication system.

A GPU Accelerated Fragment-based De Novo Ligand Design by a Bayesian Optimization Algorithm (Bioinformatics Vol.5)
Wahib Mohamed, Munawar Asim, Munetomo Masaharu
情報処理学会論文誌 論文誌トランザクション, 2011, 2, 7, 17, 情報処理学会, Apr. 2012
English

How to Apply Evolutionary Computation
MUNETOMO Masaharu, SOMEYA Hiroshi
The Journal of The Institute of Electrical Engineers of Japan, 132, 4, 204, 207, The Institute of Electrical Engineers of Japan, 01 Apr. 2012
Japanese, This article has no abstract.

-　　　　　　　　　　　　　　　
Hiroki Kashiwazaki, Masaharu Munetomo, Yoshiaki Takai
Proceedings of NORTH Internet Symposium 2012, 18, 101, 108, Feb. 2012, [Peer-reviewed]
Japanese, Symposium

Toward A Genetic Algorithm Based Flexible Approach for the Management of Virtualized Application Environments in Cloud Platforms
Omar Abdul-Rahman, Masaharu Munetomo, Kiyoshi Akama
2012 21ST INTERNATIONAL CONFERENCE ON COMPUTER COMMUNICATIONS AND NETWORKS (ICCCN), 1, 9, 2012, [Peer-reviewed]
English, International conference proceedings

Estimation of Distribution Algorithm with Mixture of Bayesian Networks
Hori Shinya, Munetomo Masaharu, Akama Kiyoshi
Transaction of the Japanese Society for Evolutionary Computation, 3, 2, 63, 72, The Japanese Society for Evolutionary Computation, 2012, [Peer-reviewed]
Japanese, This paper proposes a new method of Estimation Distribution Algorithm (EDA) named Bayesian Optimization Algorithm with Mixture Distribution (BOA-MD) that employs mixture of multiple Bayesian Networks to solve complex problems. In order to solve complex problems that are modeled by multiple Bayesian networks with hidden variables, the original BOA needs a large computation cost to model multiple probabilistic structures as a large, complex Bayesian network.The BOA-MD tries to build multiple models of Bayesian networks considering hidden variables with Expectation Maximization (EM) method to express all the structures of probabilistic distribution.The mixture of Bayesian networks is composed of a hidden variable C and some Bayesian Networks. Each composed Bayesian network can express each problem structure of multiple distributions. We perform numerical experiments by two test functions: Cross-Trap function and Triple-Trap function. These two test functions are to represent problems with multiple distributions. BOA-MD can solve these test problems with smaller number of fitness evaluations and larger modeling overheads than those by BOA for Cross-Trap5 function. This is because BOA-MD needs large computation time to construct Mixture of Bayesian Network. The BOA-MD can solve the problem faster than the original BOA when the overhands of each fitness evaluation becomes larger. At Triple-Trap function, BOA-MD can detect better solution than BOA.

Toward A Genetic Algorithm Based Flexible Approach for the Management of Virtualized Application Environments in Cloud Platforms
Omar Abdul-Rahman, Masaharu Munetomo, Kiyoshi Akama
2012 21ST INTERNATIONAL CONFERENCE ON COMPUTER COMMUNICATIONS AND NETWORKS (ICCCN), 2012, [Peer-reviewed]
English, International conference proceedings

How to Apply Evolutionary Computation
MUNETOMO MASAHARU, SOMEYA HIROSHI
電気学会誌, 132, 4, 204, 207, 2012
Japanese

A GPU accelerated fragment-based de novo ligand design by a bayesian optimization algorithm
Mohamed Wahib, Asim Munawar, Masaharu Munetomo, Kiyoshi Akama
IPSJ Transactions on Bioinformatics, 5, 7, 17, 2012, [Peer-reviewed]
English, Scientific journal

410 Study of a light-received system with configuration of a plant shoot
Nakai Shingo, Obara Shinya, Tanaka Eiichi, Konno Daisuke, Munetomo Masaharu
The Proceedings of the Symposium on Environmental Engineering, 2012, 0, 277, 278, 2012
Japanese

Development of a Plant Configuration Photoreceptor Optimized for the Objective of Maximizing Light-Received Quantity
Eiichi Tanaka, Shin'ya Obara, Shingo Nakai, Daisuke Konno, Masaharu Munetomo
2012 15TH INTERNATIONAL CONFERENCE ON ELECTRICAL MACHINES AND SYSTEMS (ICEMS 2012), 2012
English, International conference proceedings

北海道大学アカデミッククラウドにおけるコンテンツマネジメントシステムの展開
棟朝雅晴, 高井昌彰
情報科学技術フォーラム, FIT 2011, 15-18, 22 Aug. 2011
Japanese

A framework for cloud embedded Web services utilized by cloud applications
Mohamed Wahi, Asim Munawar, Masaharu Munetomo, Kiyoshi Akama
Proceedings - 2011 IEEE World Congress on Services, SERVICES 2011, 265, 271, IEEE Computer Society, 2011, [Peer-reviewed]
English, International conference proceedings

Solving extremely difficult MINLP problems using adaptive resolution Micro-GA with Tabu search
Asim Munawar, Mohamed Wahib, Masaharu Munetomo, Kiyoshi Akama
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 6683, 203, 217, Springer, 2011, [Peer-reviewed]
English, International conference proceedings

Optimization of Parallel Genetic Algorithms for nVidia GPUs
Mohamed Wahib, Asim Munawar, Masaharu Munetomo, Kiyoshi Akama
2011 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC), 803, 811, 2011, [Peer-reviewed]
English, International conference proceedings

Advanced Genetic Algorithm to solve MINLP problems over GPU
Asim Munawar, Mohamed Wahib, Masaharu Munetomo, Kiyoshi Akama
2011 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC), 318, 325, 2011, [Peer-reviewed]
English, International conference proceedings

Realizing Robust and Scalable Evolutionary Algorithms toward Exascale Era
Masaharu Munetomo
2011 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC), 312, 317, 2011, [Peer-reviewed]
English, International conference proceedings

Multi-level autonomic architecture for the management of virtualized application environments in cloud platforms
Omar Abdul-Rahman, Masaharu Munetomo, Kiyoshi Akama
Proceedings - 2011 IEEE 4th International Conference on Cloud Computing, CLOUD 2011, 754, 755, IEEE, 2011, [Peer-reviewed]
English, International conference proceedings

北海道大学アカデミッククラウドにおけるコンテンツマネジメントシステムの展開　　　　　　　　　　　　　　　
情報処理学会第10回情報科学技術フォーラム論文集（査読付論文）, RL-004, 2011

A Framework for Problem-Specific QoS Based Scheduling in Grids　　　　　　　　　　　　　　　
Advances in Grid Computing, 19, 28, 2011

An improved binary-real coded genetic algorithm for real parameter optimization
Omar Abdul-Rahman, Masaharu Munetomo, Kiyoshi Akama
Proceedings of the 2011 3rd World Congress on Nature and Biologically Inspired Computing, NaBIC 2011, 149, 156, 2011
English, International conference proceedings

An adaptive resolution hybrid binary-real coded genetic algorithm
Abdul-Rahman Omar Arif, Munetomo Masaharu, Akama Kiyoshi
Artificial Life and Robotics, 16, 1, 121, 124, 2011
English

The design, usage, and performance of GridUFO: A Grid based Unified Framework for Optimization
Munawar Asim, Wahib Mohamed, Munetomo Masaharu, Akama Kiyoshi
Future Generation Computer Systems, 26, 4, 633, 644, Apr. 2010
English

A Light Framework for the Unified Representation and Execution of Variant Tasks in a Grid Based Environment.
Mohamed Wahib, Asim Munawar, Masaharu Munetomo, Kiyoshi Akama
Proceedings of the International Conference on Parallel and Distributed Processing Techniques and Applications, PDPTA 2010, Las Vegas, Nevada, USA, July 12-15, 2010, 2 Volumes, 658, 664, CSREA Press, 2010, [Peer-reviewed]

A Proposal for Zoning Crossover of Hybrid Genetic Algorithms for Large-scale Traveling Salesman Problems
Masafumi Kuroda, Kunihito Yamamori, Masaharu Munetomo, Moritoshi Yasunaga, Ikuo Yoshihara
2010 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC), 1, 6, 2010, [Peer-reviewed]
English, International conference proceedings

Live Migration-based Resource Managers for Virtualized Environments: A Survey
Omar Abdul-Rahman, Masaharu Munetomo, Kiyoshi Akama
PROCEEDINGS OF THE FIRST INTERNATIONAL CONFERENCE ON CLOUD COMPUTING, GRIDS, AND VIRTUALIZATION (CLOUD COMPUTING 2010), 32, 40, 2010
English, International conference proceedings

A Bayesian Optimization Algorithm For De Novo Ligand Design Based Docking Running Over GPU
Mohamed Wahib, Asim Munawar, Masaharu Munetomo, Kiyoshi Akama
2010 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC), 17, 24, 2010
English, International conference proceedings

Development of a Novel Crossover of Hybrid Genetic Algorithms for Large-scale Traveling Salesman Problems　　　　　　　　　　　　　　　
Proceedings of the Fifteenth International Symposium on Artificial Life and Robotics 2010, 828, 831, 2010

Implementation and Optimization of cGA+LS to solve Capacitated VRP over Cell/B.E.
Munawar Asim, Wahib Mohamed, Munetomo Masaharu, Akama Kiyoshi
International Journal of Advancements in Computing Technology, 1, 2, 16, 28, Advanced Institute of Convergence IT, Dec. 2009
English, This paper presents a case study to illustrate the design and implementation of cellular Genetic Algorithm (cGA) with Local Search (LS) to solve Capacitated Vehicle Routing Problem (CVRP) over Cell Broadband Engine (Cell BE). Cell BE is a heterogeneous, distributed memory multicore processor architecture for multimedia applications with regular memory access requirements. It has one 64-bit Power Processing Element (PPE) that acts as the main processor and 8 Synergistic Processing Elements (SPEs) with only 256 KB of local memory, each for instructions and data. GAs on the other hand use popu...

An automated ligand evolution system using Bayesian optimization algorithm
Munetomo Masaharu, Akama Kiyoshi, Maeda Haruki
WSEAS Transactions on Information Science and Applications, 6, 5, 788, 797, World Scientific and Engineering Academy and Society, May 2009
English, Ligand docking checks whether a drug chemical called ligand matches the target receptor protein of human organ or not. Docking by computer simulation is becoming popular in drug design process to reduce cost and time of the chemical experiments. This paper presents a novel approach generating optimal ligand structures from scratch based on de novo ligand design approach employing Bayesian optimization algorithm to realize an automated design of drug and other chemical structures. The proposed approach searches an optimal structure of ligand that minimizes bond energy to the receptor protein...

Theoretical and Empirical Analysis of a GPU based Parallel Bayesian Optimization Algorithm
Asim Munawar, Mohamed Wahib, Masaharu Munetomo, Kiyoshi Akama
2009 INTERNATIONAL CONFERENCE ON PARALLEL AND DISTRIBUTED COMPUTING, APPLICATIONS AND TECHNOLOGIES (PDCAT 2009), 457, +, 2009, [Peer-reviewed]
English, International conference proceedings

Theoretical and Empirical Analysis of a GPU based Parallel Bayesian Optimization Algorithm
Asim Munawar, Mohamed Wahib, Masaharu Munetomo, Kiyoshi Akama
2009 INTERNATIONAL CONFERENCE ON PARALLEL AND DISTRIBUTED COMPUTING, APPLICATIONS AND TECHNOLOGIES (PDCAT 2009), 457, +, 2009
English, International conference proceedings

An Automated Ligand Evolution System using Bayesian Optimization Algorithm　　　　　　　　　　　　　　　
WSEA Transactions on Information Science and Applications, 6, 5, 788, 797, 2009

De Novo Ligand Evolution using Bayesian Optimization Algorithms　　　　　　　　　　　　　　　
Proceedings of the 10th WSEAS International Conference on Evolutionary Computing, 126, 131, 2009

Hybrid of genetic algorithm and local search to solve MAX-SAT problem using nVidia CUDA framework
Munawar Asim, Wahib Mohamed, Munetomo Masaharu, Akama Kiyoshi
Genetic Programming and Evolvable Machines, 10, 4, 391, 415, 2009
English

Introducing assignment functions to Bayesian optimization algorithms
Munetomo Masaharu, Murao Naoya, Akama Kiyoshi
Information Sciences, 178, 1, 152, 163, 02 Jan. 2008
English

Model for Dynamic Grain Sizing Through Compound Parallelization for an Optimization Problem Solving Grid Application
M. Wahib, Asim Munawar, Masaharu Munetomo, Akama Kiyoshi
2008 9TH IEEE/ACM INTERNATIONAL CONFERENCE ON GRID COMPUTING, 316, +, 2008
English, International conference proceedings

SOAG: Service Oriented Architectured Grids and Adoption of Application specific QoS Attributes
M. Wahib, Asim Munawar, Masaharu Munetomo, Akama Kiyoshi
2008 9TH IEEE/ACM INTERNATIONAL CONFERENCE ON GRID COMPUTING, 346, +, 2008
English, International conference proceedings

Solving Large Instances of Capacitated Vehicle Routing Problem over Cell BE
Asim Munawar, Mohamed Wahib, Masaharu Munetomo, Kiyoshi Akama
HPCC 2008: 10TH IEEE INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING AND COMMUNICATIONS, PROCEEDINGS, 131, +, 2008
English, International conference proceedings

A Survey: Genetic Algorithms and the Fast Evolving World of Parallel Computing
Asim Munawar, Mohamed Wahib, Masaharu Munetomo, Kiyoshi Akama
HPCC 2008: 10TH IEEE INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING AND COMMUNICATIONS, PROCEEDINGS, 897, +, 2008
English, International conference proceedings

A General Service-Oriented Grid Computing Framework For Global Optimization Problem Solving
M. Wahib, Asim Munawar, Masaharu Munetomo, Akama Kiyoshi
2008 IEEE INTERNATIONAL CONFERENCE ON SERVICES COMPUTING, PROCEEDINGS, VOL 2, 563, +, 2008
English, International conference proceedings

Empirical Investigations on Parallel Competent Genetic Algorithms　　　　　　　　　　　　　　　
Proceedings of the 2008 Genetic and Evolutionary Computation Conference, 1073, 1080, 2008

Linkage Identification in Evolutionary Computation(New Development in Evolutionary Computation)
MUNETOMO Masaharu
SYSTEMS, CONTROL AND INFORMATION, 52, 10, 362, 367, THE INSTITUTE OF SYSTEMS, CONTROL AND INFORMATION ENGINEERS, 2008
Japanese, 「進化計算の新展開特集号」解説

Parallel GEAs with linkage analysis over grid
Asim Munawar, Mohamed Wahib, Masaharu Munetomo, Kiyoshi Akama
Studies in Computational Intelligence, 157, 159, 187, 2008, [Peer-reviewed]
English, Scientific journal

A network design problem by a GA with linkage identification and recombination for overlapping building blocks
Miwako Tsuji, Masaharu Munetomo, Kiyoshi Akama
Studies in Computational Intelligence, 157, 441, 459, 2008, [Peer-reviewed]
English, Scientific journal

Linkage analysis in genetic algorithms
Miwako Tsuji, Masaharu Munetomo
Studies in Computational Intelligence, 137, 251, 279, 2008, [Peer-reviewed]
English, Scientific journal

複雑なビルディングブロック重複を持つ問題に対する交叉手法の提案
辻 美和子, 棟朝 雅晴, 赤間 清
情報処理学会論文誌. 数理モデル化と応用, 48, 15, 23, 33, 一般社団法人情報処理学会, 15 Oct. 2007
Japanese, 遺伝的アルゴリズムによる効率的な探索のために,同一のビルディングブロック(building block, BB)を構成する遺伝子座の集合を検出する手法は多く提案されている(Heckendornら).しかしながら,これらの手法から得られたリンケージ情報を利用して効果的に交叉を行う方法については,十分な検討がなされてこなかった.特に重複するBBを持つ問題ではYuら(2005)の交叉手法のみが知られている.しかし,彼らの手法はBBの重複構造が複雑になったとき,頻繁にBBを破壊し,かつ十分な交叉パターンが得られないために,効率的に機能しない.本論文では,Yuらの手法を拡張し,BB破壊をできるだけ抑えながら,新たな異なる探索点を与える交叉手法を提案する.提案される手法は,コンテクスト依存交叉(Context Dependent Crossover, CDC)と呼ばれ,与えられた親個体組の値を調査したうえで,交換する遺伝子座を決定する.CDCは,リンケージ同定手法と併用されることで,重複するBBを持つ問題を探索する強力なアルゴリズムを提供する.また,提案手法の性能を確認するために,重複の複雑さが制御可能なテスト関数を設計する.

MHGrid: Towards an ideal optimization environment for global optimization problems using Grid computing
M. Wahib, Asim Munawar, Masaharu Munetomo, Kiyoshi Akama
EIGHTH INTERNATIONAL CONFERENCE ON PARALLEL AND DISTRIBUTED COMPUTING, APPLICATIONS AND TECHNOLOGIES, PROCEEDINGS, 167, +, 2007
English, International conference proceedings

複雑なビルディングブロック重複を持つ問題に対する交叉手法の提案　　　　　　　　　　　　　　　
情報処理学会論文誌「数理モデル化と応用」, 48, SIG15, 23, 33, 2007

Optimization problem solving framework employing GAs with linkage identification over a Grid environment
Asim Munawar, Masaharu Munetomo, Kiyoshi Akama
2007 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1-10, PROCEEDINGS, 1191, +, 2007
English, International conference proceedings

A network design problem by a GA with linkage identification and recombination for overlapping building blocks
Miwako Tsuji, Masaharu Munetomo, Kiyoshi Akama
2007 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1-10, PROCEEDINGS, 349, 356, 2007
English, International conference proceedings

Standardization of Interfaces for Meta-Heuristics based Problem Solving Framework over Grid Environment　　　　　　　　　　　　　　　
Proceedings of the HPC Asia 2007, 129, 136, 2007

An intelligent scatter with estimation of distribution for tabu search
Masaharu Munetomo, Yuta Satake, Kiyoshi Akama
COMPUTER AIDED SYSTEMS THEORY- EUROCAST 2007, 4739, 465, +, 2007
English, International conference proceedings

A framework of GRID problem-solving environment employing robust evolutionary search
Masaharu Munetomo, Asim Munawar, Kiyoshi Akama
COMPUTER AIDED SYSTEMS THEORY- EUROCAST 2007, 4739, 473, +, 2007
English, International conference proceedings

On Hybridization of Bayesian Optimization and Tabu Search　　　　　　　　　　　　　　　
Proceedings of the Seventh Metaheuristics International Conference (CD-ROM), 52, 2007

Genetic algorithm to optimize fitness function with sampling error and its application to financial optimization problem
Masaru Tezuka, Masaharu Munetomo, Kiyoshi Akama
Studies in Computational Intelligence, 51, 417, 434, 2007, [Peer-reviewed]
English, Scientific journal

Linkage identification by fitness difference clustering
Miwako Tsuji, Masaharu Munetomo, Kiyoshi Akama
Evolutionary Computation, 14, 4, 383, 409, Dec. 2006
English, Scientific journal

Linkage identification by fitness difference clustering
Miwako Tsuji, Masaharu Munetomo, Kiyoshi Akama
Evolutionary Computation, 14, 4, 383, 409, 4, Dec. 2006, [Peer-reviewed]
English, Scientific journal

Linkage Identification for Real-coded Genetic Algorithms Based on Additive Decomposability and Difference Signature Independency of Objective Function
TEZUKA MASARU, MUNETOMO MASAHARU, AKAMA KIYOSHI
情報処理学会論文誌数理モデル化と応用（TOM）, 47, 14, 43, 53, Information Processing Society of Japan (IPSJ), 15 Oct. 2006
Japanese, In the case that a problem is decomposable to a number of sub-problems which can be optimized independently, the problem is solved effectively by optimizing sub-problems separately. In optimization problems by means of genetic algorithms, a set of loci of which each sub-problem consists is called linkage group. Linkage identification is the method which recognizes linkage groups. In this paper, we define the linkage of Real-Coded GAs clearly. Then we propose two new linkage identification methods, LINC-R and LIDI-R, directly based on the definition. LINC-R is based on additive decomposability and LIDI-R is based on independency of the signature of difference of an objective function. These methods effectively identify linkages.

Linkage Identification for Real-coded Genetic Algorithms Based on Additive Decomposability and Difference Signature Independency of Objective Function
TEZUKA MASARU, MUNETOMO MASAHARU, AKAMA KIYOSHI
情報処理学会論文誌. 数理モデル化と応用, 47, 14, 43, 53, Information Processing Society of Japan (IPSJ), 15 Oct. 2006
Japanese, In the case that a problem is decomposable to a number of sub-problems which can be optimized independently, the problem is solved effectively by optimizing sub-problems separately. In optimization problems by means of genetic algorithms, a set of loci of which each sub-problem consists is called linkage group. Linkage identification is the method which recognizes linkage groups. In this paper, we define the linkage of Real-Coded GAs clearly. Then we propose two new linkage identification methods, LINC-R and LIDI-R, directly based on the definition. LINC-R is based on additive decomposabili...

Risk Visualization and Decision Support for Supply Planning under Uncertain Demand
TEZUKA MASARU, HIJI MASAHIRO, MUNETOMO MASAHARU, AKAMA KIYOSHI
IPSJ journal, 47, 3, 701, 710, Information Processing Society of Japan (IPSJ), 15 Mar. 2006
Japanese, In recent years, the life-cycle of high-tech products such as personal computers and cellular phones is getting shortened. Shorter life-cycle makes the risk of opportunity loss and dead-stock disposal loss higher. Supply plan designates the quantity and the delivery date of products to supply. The plan is created based on forecasted demand in order to cover the demand. Statistical method is used to forecast, however, shorter life-cycle also makes it difficult to obtain enough amounts of historical record to conduct statistical analysis. Consequently, the forecast accuracy declines. Since the decision made on the supply plan has a greater impact on the profitability these days, the decision based on quantitative evaluation is essential to manufacturers and resellers. In this paper, the model which explains how the profit is produced from the demand and supply is constructed. Taking advantage of the model, gross profit, opportunity loss, and dead-stock disposal loss are analyzed with Monte-Carlo method and evaluated quantitatively. The application of the proposed system on two cases shows the effectiveness of the system.

Risk Visualization and Decision Support for Supply Planning under Uncertain Demand(Information Systems for Society and Humans,Information Systems for New Application Domains)
TEZUKA MASARU, HIJI MASAHIRO, MUNETOMO MASAHARU, AKAMA KIYOSHI
IPSJ Journal, 47, 3, 701, 710, Information Processing Society of Japan (IPSJ), 15 Mar. 2006
Japanese, In recent years, the life-cycle of high-tech products such as personal computers and cellular phones is getting shortened. Shorter life-cycle makes the risk of opportunity loss and dead-stock disposal loss higher. Supply plan designates the quantity and the delivery date of products to supply. The plan is created based on forecasted demand in order to cover the demand. Statistical method is used to forecast, however, shorter life-cycle also makes it difficult to obtain enough amounts of historical record to conduct statistical analysis. Consequently, the forecast accuracy declines. Since th...

Realizing Virtual Innovative Laboratory with Robust Evolutionary Algorithms over the GRID computing system　　　　　　　　　　　　　　　
Proceedings of the 6th International Conference on Recent Advance in Soft Computing, 42, 47, 2006

A crossover for complex building blocks overlapping
Miwako Tsuji, Masaharu Munetomo, Kiyoshi Akama
GECCO 2006: GENETIC AND EVOLUTIONARY COMPUTATION CONFERENCE, VOL 1 AND 2, 1337, +, 2006
English, International conference proceedings

Control the Number of Samples to Estimate Fitness from the Perspective of Takeover Time and Optimization of Financial Criteria　　　　　　　　　　　　　　　
Proceedings of the 2006 IEEE Congress on Evolutionary Computation, 388, 394, 2006

Enhancing model-building efficiency in extended compact genetic algorithms
Masaharu Munetomo, Yuta Satake, Kiyoshi Akama
2006 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN, AND CYBERNETICS, VOLS 1-6, PROCEEDINGS, 2362, +, 2006
English, International conference proceedings

Theoretical and empirical investigations on difficulty in structure learning by estimation of distribution algorithms
Miwako Tsuji, Masaharu Munetomo, Kiyoshi Akama
2006 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN, AND CYBERNETICS, VOLS 1-6, PROCEEDINGS, 209, +, 2006
English, International conference proceedings

Genetic algorithm to optimize fitness function with sampling error and its application to financial optimization problem
Masaru Tezuka, Masaharu Munetomo, Kiyoshi Akama, Masahiro Hiji
2006 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1-6, 81, 87, 2006, [Peer-reviewed]
English, International conference proceedings

Miwako Tsuji, Masaharu Munetomo, Kiyoshi Akama: "Population Sizing of Dependency Detection by Fitness Difference Classification", Foundations of Genetic Algorithms - FOGA2005, Lecture Notes in Computer Science 3469:282-299 (2005)*　　　　　　　　　　　　　　　
2005

Miwako Tsuji, Masaharu Munetomo, Kiyoshi Akama: "Linkage Identification for Real-Values Loci by Fitness Difference Classification", Proceedings of 2005 Congress on Evolutionary Computation, 2:1317-1324 (2005)*　　　　　　　　　　　　　　　
2005

Masaharu Munetomo, Naoya Murao, Kiyoshi Akama: "Empirical Studies on Parallel Network Construction of Bayesian Optimization Algorithms", Proceedings of 2005 Congress on Evolutionary Computation, 2:1524-1531 (2005)*　　　　　　　　　　　　　　　
2005

Empirical studies on parallel network construction of Bayesian optimization algorithms
M Munetomo, N Murao, K Akama
2005 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1-3, PROCEEDINGS, 1524, 1531, 2005, [Peer-reviewed]
English, International conference proceedings

Linkage identification for real-valued loci by fitness difference classification
M Tsuji, M Munetomo, K Akama
2005 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1-3, PROCEEDINGS, 1317, 1324, 2005, [Peer-reviewed]
English, International conference proceedings

Population sizing of dependency detection by fitness difference classification
M Tsuji, M Munetomo, K Akama
FOUNDATIONS OF GENETIC ALGORITHMS, 3469, 282, 299, 2005, [Peer-reviewed]
English, Scientific journal

Multi-objective Real-coded Genetic Algorithm Approach for Supply Planning Under Uncertainty
TEZUKA MASARU, HIJI MASAHIRO, MUNETOMO MASAHARU, AKAMA KIYOSHI
IPSJ journal, 45, 10, 2287, 2296, Information Processing Society of Japan (IPSJ), 15 Oct. 2004
Japanese, Supply is planned to meet the future forecast. However, uncertainty is involeved in the supply plan since it is difficult to forecast the future demand accurately. The impact to business caused by the gap between the forecast and actual demand is called risk. Thus, supply planning methods which can maximize profit and minimize risk simultaneously is desired. The conventional method based on safety stock or buffer stock has been widely used, whose main purpose is to prevent the occurrence of opportunity loss. In order to simulate the uncertainty and evaluate the profit and risk, we introduced Monte Carlo simulation. According to the fitness calculated by the simulation, a genetic algorithm optimizes the profit, risk, opportunity loss, and final inventory quantity of supply planning problems. The approach was tested on the supply planning data and has achieved a remarkable result.

Multi-objective Real-coded Genetic Algorithm Approach for Supply Planning Under Uncertainty(Applications of Information Systems to Society and Interprises)
TEZUKA MASARU, HIJI MASAHIRO, MUNETOMO MASAHARU, AKAMA KIYOSHI
IPSJ Journal, 45, 10, 2287, 2296, Information Processing Society of Japan (IPSJ), 15 Oct. 2004
Japanese, Supply is planned to meet the future forecast. However, uncertainty is involeved in the supply plan since it is difficult to forecast the future demand accurately. The impact to business caused by the gap between the forecast and actual demand is called risk. Thus, supply planning methods which can maximize profit and minimize risk simultaneously is desired. The conventional method based on safety stock or buffer stock has been widely used, whose main purpose is to prevent the occurrence of opportunity loss. In order to simulate the uncertainty and evaluate the profit and risk, we introduce...

Linkage Identification for Problems with Hierarchical Structure
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI
情報処理学会論文誌数理モデル化と応用（TOM）, 45, 2, 22, 31, Information Processing Society of Japan (IPSJ), 15 Feb. 2004
Japanese, To avoid building block destructions, linkage identification techniques are proposed, which tries to identify a set of loci tightly-linked explicitly before performing genetic optimizations. Real-world problems, especially large-scaled complex problems, sometimes take hierarchical structures in which building blocks have recursive interdependencies. Existing linkage identification algorithms only consider interactions between loci in a same building block and assume no interdependency between building blocks. In this paper, the LIEM2 (Linkage Identification with Epistasis Measure considering Monotonicity) - a single layer linkage identification algorithm based on non-monotonicity conditions - is extended to identify hierarchical multi-layered linkage groups in order to search more accurate structures of real-world problems. The hierarchical linkage identification identifies linkage groups hierarchically in a recursive manner employing niching which preserve various building block candidates.

Linkage Identification for Problems with Hierarchical Structure (Theory)(Evolutionary Computation)
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI
情報処理学会論文誌. 数理モデル化と応用, 45, 2, 22, 31, Information Processing Society of Japan (IPSJ), 15 Feb. 2004
Japanese, To avoid building block destructions, linkage identification techniques are proposed, which tries to identify a set of loci tightly-linked explicitly before performing genetic optimizations. Real-world problems, especially large-scaled complex problems, sometimes take hierarchical structures in which building blocks have recursive interdependencies. Existing linkage identification algorithms only consider interactions between loci in a same building block and assume no interdependency between building blocks. In this paper, the LIEM2 (Linkage Identification with Epistasis Measure considerin...

Masaru Tezuka, Masaharu Munetomo, Kiyoshi Akama: Selection Efficiency and Sampling Error on Genetic Algorithms Optimization under Uncertainty, Proceedings of 2004 Simulated Evolution and Learning (SEAL2004), (CD-ROM) (2004)*　　　　　　　　　　　　　　　
2004

Masaharu Munetomo, Naoya Murao, Kiyoshi Akama: "Empirical Investigations on Parallelized Linkage Identification", Parallel Problem Solving from Nature - PPSN VIII, Lecture Notes in Computer Science, 3242:322-331 (2004)*　　　　　　　　　　　　　　　
2004

Naoya Murao, Masaharu Munetomo, Kiyoshi Akama: "Performance Comparison between Parallel GA Based on Linkage Identification and Parallel Bayesian Optimization Algorithm", Proceedings of the International Conference on Cybernetics and Information Technol・・・　　　　　　　　　　　　　　　
2004
Naoya Murao, Masaharu Munetomo, Kiyoshi Akama: "Performance Comparison between Parallel GA Based on Linkage Identification and Parallel Bayesian Optimization Algorithm", Proceedings of the International Conference on Cybernetics and Information Technologies, Systems and Applications (CITSA2004), 3:136-141 (2004)*

Masaru Tezuka, Masaharu Munetomo, Kiyoshi Akama: "Linkage Identification by Nonlinearity Check for Real-coded Genetic Algorithms", Genetic and Evolutionary Computation - GECCO2004 Part 2, Lecture Notes in Computer Science, 3103:222-233 (2004)*　　　　　　　　　　　　　　　
2004

Miwako Tsuji, Masaharu Munetomo, Kiyoshi Akama: "Modeling Dependencies of Loci with String Classification According to Fitness Differences", Genetic and Evolutionary Computation - GECCO2004 Part 2, Lecture Notes in Computer Science, 3103:246-257 (2004)*　　　　　　　　　　　　　　　
2004

Designing a Distributed Algorithm for Bandwidth Allocation with a Genetic Algorithm
Hidehiro Kobayashi, Masaharu Munetomo, Kiyoshi Akama, Yoshiharu Sato
Systems and Computers in Japan, 35, 3, 37, 45, John Wiley and Sons Inc., 2004
English, Scientific journal

Masaharu Munetomo: "Estimation of Distribution Algorithms without Explicit Selections", Proceedings of The 8th World Multi-Conference on Systemics, Cybernetics and Informatics (SCI2004), 5:80-85 (2004)　　　　　　　　　　　　　　　
2004

Estimation of distribution algorithms without explicit selections
M Munetomo
8TH WORLD MULTI-CONFERENCE ON SYSTEMICS, CYBERNETICS AND INFORMATICS, VOL V, PROCEEDINGS, 80, 85, 2004, [Peer-reviewed]
English, International conference proceedings

Modeling dependencies of loci with string classification according to fitness differences
M Tsuji, M Munetomo, K Akama
GENETIC AND EVOLUTIONARY COMPUTATION GECCO 2004 , PT 2, PROCEEDINGS, 3103, 246, 257, 2004, [Peer-reviewed]
English, Scientific journal

Empirical investigations on parallelized linkage indentification
M Munetomo, N Murao, K Akama
PARALLEL PROBLEM SOLVING FROM NATURE - PPSN VIII, 3242, 322, 331, 2004, [Peer-reviewed]
English, Scientific journal

Performance comparison between parallel GA based on linkage identification and parallel Bayesian optimization algorithm
N Murao, M Munetomo, K Akama
ISAS/CITSA 2004: International Conference on Cybernetics and Information Technologies, Systems and Applications and 10th International Conference on Information Systems Analysis and Synthesis, Vol 3, Proceedings, 136, 141, 2004, [Peer-reviewed]
English, International conference proceedings

Linkage identification by nonlinearity check for real-coded genetic algorithms
M Tezuka, M Munetomo, K Akama
GENETIC AND EVOLUTIONARY COMPUTATION GECCO 2004 , PT 2, PROCEEDINGS, 3103, 222, 233, 2004, [Peer-reviewed]
English, Scientific journal

Miwako Tsuji, Masaharu Munetomo, and Kiyoshi Akama: "Metropolitan Area Network Design Using GA Based on Hierarchical Linkage Identification", Genetic and Evolutionary Computation Part 2, Lecture Notes in Computer Science 2724:1616-1617 (2003)*　　　　　　　　　　　　　　　
2003

Masaharu Munetomo, Naoya Murao, and Kiyoshi Akama: "A Parallel Genetic Algorithm Based on Linkage Identification", Genetic and Evolutionary Computation Part 1, Lecture Notes in Computer Science 2723:1222-1233 (2003)*　　　　　　　　　　　　　　　
2003

Metropolitan area network design using GA based on Hierarchical linkage identification
Miwako Tsuji, Masaharu Munetomo, Kiyoshi Akama
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2724, 1614, 1615, Springer Verlag, 2003, [Peer-reviewed]
English, International conference proceedings

Metropolitan area network design using GA based on hierarchical linkage identification
M Tsuji, M Munetomo, K Akama
GENETIC AND EVOLUTIONARY COMPUTATION - GECCO 2003, PT II, PROCEEDINGS, 2724, 1616, 1617, 2003, [Peer-reviewed]
English, Scientific journal

A parallel genetic algorithm based on linkage identification
M Munetomo, N Murao, K Akama
GENETIC AND EVOLUTIONARY COMPUTATION - GECCO 2003, PT I, PROCEEDINGS, 2723, 1222, 1233, 2003, [Peer-reviewed]
English, Scientific journal

Proposal of a Linkage Identification Method Based on Epistasis Measure
MUNETOMO MASAHARU
情報処理学会論文誌数理モデル化と応用（TOM）, 43, 10, 6, 13, Information Processing Society of Japan (IPSJ), 15 Nov. 2002
Japanese, Genetic Algorithms realize effective search by exchanging building blocks through genetic recombinations. To realize effective genetic search, linkage identification becomes important which detects a set of loci tightly linked to form a building block. Several methods are already proposed to identify linkage such as linkage learning algorithms based on probabilistic models and linkage identification procedures based on nonlinearity or non-monotonicity detection. In this paper, we extend the Linkage Identification by Nonlinearity Check (LINC) which identifies linkage based on nonlinearity detection by defining an epistasis measure for each pair of loci to realize linkage identification with the epistasis measure.

Proposal of a Linkage Identification Method Based on Epistasis Measure(≤Special Issue≥Special Issue on Evolutionary Computation)
MUNETOMO MASAHARU
情報処理学会論文誌. 数理モデル化と応用, 43, 10, 6, 13, Information Processing Society of Japan (IPSJ), 15 Nov. 2002
Japanese, Genetic Algorithms realize effective search by exchanging building blocks through genetic recombinations. To realize effective genetic search, linkage identification becomes important which detects a set of loci tightly linked to form a building block. Several methods are already proposed to identify linkage such as linkage learning algorithms based on probabilistic models and linkage identification procedures based on nonlinearity or non-monotonicity detection. In this paper, we extend the Linkage Identification by Nonlinearity Check (LINC) which identifies linkage based on nonlinearity de...

A Genetic Algorithm Using Linkage Identification for Metropolitan Area Network Design
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI
情報処理学会論文誌数理モデル化と応用（TOM）, 43, 7, 80, 91, Information Processing Society of Japan (IPSJ), 15 Sep. 2002
Japanese, In genetic algorithms, it is important to encode strings ensuring tight linkage for their building blocks. In network design problems, however, it is difficult to encode strings appropriately because network design is dependent not only on geographical constraints but also on other complex factors such as bias on traffic demands, routing policy, and so on. Although there's many applications of genetic algorithms to network topology design, most of them haven't paid attention to tight encoding of building blocks, or considered only geographical characteristics. In order to realize tight linkage among loci and realize effective genetic search, this paper introduces LIEM (Linkage Identification with Epistasis Measure) - a technique for identifying linkage sets, sets of loci tightly liked to form building blocks - to realize effective network design. Through empirical studies, we show the effectiveness of the network design with the LIEM compared to that with conventional genetic algorithms.

A Genetic Algorithm Using Linkage Identification for Metropolitan Area Network Design
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI
情報処理学会論文誌. 数理モデル化と応用, 43, 7, 80, 91, Information Processing Society of Japan (IPSJ), 15 Sep. 2002
Japanese, In genetic algorithms, it is important to encode strings ensuring tight linkage for their building blocks. In network design problems, however, it is difficult to encode strings appropriately because network design is dependent not only on geographical constraints but also on other complex factors such as bias on traffic demands, routing policy, and so on. Although there's many applications of genetic algorithms to network topology design, most of them haven't paid attention to tight encoding of building blocks, or considered only geographical characteristics. In order to realize tight link...

Load Balancing Routing with Genetic Algorithm Based on Link Load Metric
YAMAGUCHI Naohiko, MUNETOMO Masaharu, AKAMA Kiyoshi, SATO Yoshiharu
Transactions of Information Processing Society of Japan, 43, 7, 2359, 2367, Information Processing Society of Japan (IPSJ), 15 Jul. 2002
Japanese, In packet switching networks such as the Internet, to utilize network resources effectively, routing algorithms with genetic algorithms have been proposed which generate alternative routes by genetic operators to balance loads among them and prevent congestions. This paper proposes adaptive genetic operators based on link load metric for the genetic routing algorithms in order to realize rapid evaluations of network load status and effective generations of alternative routes. Through simulation experiments performed on a network simulator, we show the effectiveness of the proposed method.

Load Balancing Routing with Genetic Algorithm Based on Link Load Metric
YAMAGUCHI NAOHIKO, MUNETOMO MASAHARU, AKAMA KIYOSHI, SATO YOSHIHARU
IPSJ Journal, 43, 7, 2359, 2367, Information Processing Society of Japan (IPSJ), 15 Jul. 2002
Japanese, In packet switching networks such as the Internet, to utilize network resources effectively, routing algorithms with genetic algorithms have been proposed which generate alternative routes by genetic operators to balance loads among them and prevent congestions. This paper proposes adaptive genetic operators based on link load metric for the genetic routing algorithms in order to realize rapid evaluations of network load status and effective generations of alternative routes. Through simulation experiments performed on a network simulator, we show the effectiveness of the proposed method.

Designing a Distributed Algorithm for Bandwidth Allocation with Genetic Algorithms
KOBAYASHI Hidehiro, MUNETOMO Masaharu, AKAMA Kiyoshi, SATO Yoshiharu
The Transactions of the Institute of Electronics,Information and Communication Engineers., 85, 5, 445, 452, The Institute of Electronics, Information and Communication Engineers, 01 May 2002
Japanese, 既存のネットワーク上で一定時間,一定の帯域幅を必要とするストリーム通信を実現する場合,ユーザからのサービス要求を満たすにはネットワーク資源を効率的に配分する帯域幅割当てアルゴリズムが重要となる.小規模なネットワークにおける割当ての最適化は,あるノードがネットワーク情報を集中的に管理し一括した割当てを実行する方法が最も効率が良い.しかし,帯域幅割当てはネットワーク規模の増加に比例して最適化に必要とする計算時間が増加するため,実装を考慮した場合にはアルゴリズムの分散化が必要となる.また分散化することによりネットワーク障害が発生した場合でも割当てが可能となり,アルゴリズムの耐障害性が向上する.本論文では,既存の遺伝的アルゴリズムによる帯域幅割当てアルゴリズムの分散化を目的とし,同時にネットワーク障害が発生した場合に全通信の再割当てを行うことで障害への対応を実現した改良アルゴリズムを提案する.本アルゴリズムの特徴は,計算時間を軽減するために各ノードで割当ての最適化を実行する局所的最適化と,全ノードで局所的最適化を実行しながら勝ち抜き戦を行うことでネットワーク全体を最適化する大域的最適化とを組み合わせて構成していることである.更にシミュレーション実験を行うことで,集中型アルゴリズムと提案アルゴリズムの性能について検討する.

「遺伝的アルゴリズムによる帯域幅割当てのための分散アルゴリズムの設計」　　　　　　　　　　　　　　　
『電子情報通信学会論文誌 D-I』, J85-D-I, 5, 445, 452, 2002

Yuichi Yamamoto, Takahiko Ishikawa, Kiyoshi Akama, Masaharu Munetomo: "A Foundation for Algorithm Generation by Transforming Meta-descriptions", Proceedings of the 2002 International Conference on Fuzzy Systems and Knowledge Discovery, 12-716 (2002)*　　　　　　　　　　　　　　　
2002

Masaharu Munetomo, Miwako Tsuji, Kiyoshi Akama: "Metropolitan Area Network Design Using GA Based on Linkage Identification with Epistasis Measures", Proceedings of the 4th Asia-Pacific Conference on Simulated Evolution and Learning, 652-656 (2002)*　　　　　　　　　　　　　　　
2002

Masaharu Munetomo: "Linkage Identification with Epistasis Measure Considering Monotonicity Conditions", Proceedings of the 4th Asia-Pacific Conference on Simulated Evolution and Learning, 550-554 (2002)*　　　　　　　　　　　　　　　
2002

Masaharu Munetomo: "Linkage Identification Based on Epistasis Measures to Realize Efficient Genetic Algorithms", Proceedings of the 2002 Congress on Evolutionary Computation, 1332-1337 (2002)*　　　　　　　　　　　　　　　
2002

Linkage identification based on epistasis measures to realize efficient genetic algorithms
M Munetomo
CEC'02: PROCEEDINGS OF THE 2002 CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1 AND 2, 1332, 1337, 2002, [Peer-reviewed]
English, International conference proceedings

Hidehiro Kobayashi, Masaharu Munetomo, Kiyoshi Akama, and Yoshiharu Sato: A Distributed Algorithm for Bandwidth Allocation in Multimedia Networks, Proceedings of the 5th International Conference on Artificial Evolution, 251-262 (2001)*　　　　　　　　　　　　　　　
2001

Masaharu Munetomo, Naohiko Yamaguchi, Kiyoshi Akama, and Yoshiharu Sato: Empirical Investigations oon the Genetic Adaptive Routing Algorithm in the Internet, Proceedings of the Congress on Evolutionary Computation 2001, 1236-1243 (2001)*　　　　　　　　　　　　　　　
2001

Masaharu Munetomo: The Genetic Adaptive Routing Algorithm, in Telecommunications Optimisation: Heuristic and Adaptive Methods, pp.151-166, John Weily & Sons*　　　　　　　　　　　　　　　
2001

Empirical investigations on the genetic adaptive routing algorithm in the Internet
M Munetomo, N Yamaguchi, K Akama, Y Sato
PROCEEDINGS OF THE 2001 CONGRESS ON EVOLUTIONARY COMPUTATION, VOLS 1 AND 2, 1236, 1243, 2001, [Peer-reviewed]
English, International conference proceedings

Masaharu Munetomo: The Genetic Adaptive Routing Algorithm, in Telecommunications Optimisation: Heuristic and Adaptive Methods, pp.151-166, John Weily & Sons (2000). *　　　　　　　　　　　　　　　
2000

Masaharu Munetomo: Network Routing with the Use of Evolutionary Methods, in Computational Intelligence in Telecommunication Networks (分担 : Witold Pedrycz and Athanasios V. Vasilakos, editors), CRC Press (2000).　　　　　　　　　　　　　　　
2000

Masaharu Munetomo, David E. Goldberg: Linkage Identification by Non-monotonicity Detection for Overlapping functions, Evolutionary Computation, vol.7, No.4, pp.377-398*　　　　　　　　　　　　　　　
1999

Masaharu Munetomo, David E. Goldberg: A Genetic Algorithm Using Linkage Identification by Nonlinearity Check, Proceedings of the 1999 IEEE Conference on System, Man, and Cybernetics (1999)　　　　　　　　　　　　　　　
1999

Masaharu Munetomo, David E. Goldberg: Identifying Linkage Groups by Nonlinearity/Non-monotonicity Detection, Proceedings of the Genetic and Evolutionary Computation Conference (GECCO-99), pp.433-440 (1999)　　　　　　　　　　　　　　　
1999

Masaharu Munetomo, David E. Goldberg: Linkage Identification by Non-monotonicity Detection for Overlapping functions, Evolutionary Computation, vol.7, No.4, pp.377-398 (1999)　　　　　　　　　　　　　　　
1999

Linkage Identification by Non-monotonicity Detection for Overlapping Functions
Munetomo Masaharu, Goldberg David E
Evolutionary Computation, 7, 4, 377, 398, MIT Press, 1999
English, This paper presents the linkage identification by non-monotonicity detection (LIMD) procedure and its extension for overlapping functions by introducing the tightness detection (TD) procedure. The LIMD identifies linkage groups directly by performing order-2 simultaneous perturbations on a pair of loci to detect monotonicity/non-monotonicity of fitness changes. The LIMD can identify linkage groups with at most order of k when it is applied to O(2k) strings. The TD procedure calculates tightness of linkage between a pair of loci based on the linkage groups obtained by the LIMD. By removing l...

A Migration Scheme of the Genetic Adaptive Routing Algorithm　　　　　　　　　　　　　　　
Masaharu Munetomo, Yoshiaki Takai, Yoshiharu Sato
Proc. of IEEE Int. Conf. on Systems, Man and Cybernetics,, 2774, 2779, Jul. 1998, [Peer-reviewed]
English, International conference proceedings

An Adaptive Routing Algorithm with Load Balancing by a Genetic Algorithm
MUNEASA Masaharu, TAKAI Yoshiaki, SATO Yoshiharu
Transactions of Information Processing Society of Japan, 39, 2, 219, 227, Information Processing Society of Japan (IPSJ), 15 Feb. 1998
Japanese, This paper presents an adaptive routing algorithm which has a load balancing mechanism among alternative paths by a genetic algorithm. Conventional routing algorithms such as RIP and SPF broadcast information on routing tables or link status in a network, which yields much communication overhead and degrades total performance when the network becomes large. Conventional routing algorithms only generate the shortest path to send a packet, even if some good alternative paths are available. Our routing algorithm generates alternative paths and observes communication latency only for paths frequently used. This mechanism greatly reduces communication cost for information exchangind of the routing. Moreover this algorithm realizes load balancing on the alternative paths by distributing packets among them. We perform simulation experiments using a discrete event simulator of network communication. The result of the experiments shows that our algorithm achieves effective routing with less communication overhead.

An Adaptive Routing Algorithm with Load Balancing by a Genetic Algorithm (Multimedia・Distirbuted and Cooperative Computing)
MUNETOMO MASAHARU, TAKAI YOSHIAKI, SATO YOSHIHARU
IPSJ Journal, 39, 2, 219, 227, Information Processing Society of Japan (IPSJ), 15 Feb. 1998
Japanese, This paper presents an adaptive routing algorithm which has a load balancing mechanism among alternative paths by a genetic algorithm. Conventional routing algorithms such as RIP and SPF broadcast information on routing tables or link status in a network, which yields much communication overhead and degrades total performance when the network becomes large. Conventional routing algorithms only generate the shortest path to send a packet, even if some good alternative paths are available. Our routing algorithm generates alternative paths and observes communication latency only for paths freq...

Migration scheme for the genetic adaptive routing algorithm　　　　　　　　　　　　　　　
Masaharu Munetomo, Yoshiaki Takai, Yoshiharu Sato
Proceedings of the IEEE International Conference on Systems, Man and Cybernetics, 3, 2774, 2779, IEEE, 1998
English, International conference proceedings

遺伝的アルゴリズムによる負荷分散機構を有する適応型ルーティング　　　　　　　　　　　　　　　
情報処理学会論文誌, Vol.38, No.2, 219, 227, 1998

Shuichi Moriguti, Masaharu Munetomo, Yoshiharu Sato: A Moving Average Method for Predicting Process Resource Usage Based on a State Transition Model, Proceedings of the 1998 Conference of the North American Fuzzy Information Processing Society, pp.82-85*　　　　　　　　　　　　　　　
1998

Masaharu Munetomo, Yoshiaki Takai, Yoshiharu Sato: A Migration Scheme for the Genetic Adaptive Routing Algorithm, Proceedings of the 1998 IEEE Conference on Systems, Man, and Cybernetics, pp.2774-2779*　　　　　　　　　　　　　　　
1998

Shuichi Moriguti,Masaharu Munetomo,Yoshiharu Sato:A Moving Average Method for Predicting Process Resource Usage Based on a State Transition Model(North American Fuzzy Information Processing Society(NAFIPS'98),1998)　　　　　　　　　　　　　　　
1998

Masaharu Munetomo,Yoshiaki Takai,Yoshiharu Sato:A Migration Scheme for the Genetic Adaptive Routing Algorithm (IEEE Conference on Systems,Man,and Cybernetics(SMC'98),1998)　　　　　　　　　　　　　　　
1998

A moving average method for predicting process resource usage based on a state transition model
S Moriguchi, M Munetomo, Y Sato
1998 CONFERENCE OF THE NORTH AMERICAN FUZZY INFORMATION PROCESSING SOCIETY - NAFIPS, 82, 85, 1998, [Peer-reviewed]
English, International conference proceedings

An Intelligent Network Routing Algorithm by a Genetic Algorithm　　　　　　　　　　　　　　　
Masaharu Munetomo, Yoshiaki Takai, Yoshiharu Sato
Proc. of the 4th Int. Conf. on Neural Information Processing, 1, 547, 550, Nov. 1997, [Peer-reviewed]
English, International conference proceedings

An Adaptive Network Routing Algorithm Employing Path Genetic Operators　　　　　　　　　　　　　　　
Masaharu Munetomo, Yoshiaki Takai, Yoshiharu Sato
Proc. of the 7th Int. Conf. on Genetic Algorithms, 643, 649, Jul. 1997, [Peer-reviewed]
English, International conference proceedings

An Application of a Stochastic Algorithmo to Invisible Matrices Games
TOMIKAWA Yuuki, MUNETOMO Masaharu, TAKAI Yoshiaki
The Transactions of the Institute of Electronics,Information and Communication Engineers., 80, 2, 700, 702, The Institute of Electronics, Information and Communication Engineers, 25 Feb. 1997
Japanese, 強化学習機構を有する遺伝的アルゴリズム (StGA) [1], 利得行列の内容が不可視であり, 可能な行動の数が多いゲームヘ適用する. StGAのもととなった確率学習オートマトンとの比較実験を通し, こうしたゲームに対するStGA適用の有効性を検証する.

「利得行列が不可視である行列ゲームへのStGAの適用」　　　　　　　　　　　　　　　
『電子情報通信学会論文誌』, J80-D-II, 2, 700, 702, 1997

Munetomo, M., Takai, Y. and Sato, Y. : "An Adaptive Network Routing Algorithm Employing Path Genetic Operators", Proceedings of the Seventh International Conference on Genetic Algorithms, 643-649 (1997)＊　　　　　　　　　　　　　　　
1997

Munetomo, M., Takai, Y. and Sato, Y. : "An Intelligent Network Routing Algorithm by a Genetic Algorithm", Proceedings of the 1997 International Conference on Neural Information Processing and Intelligent Information Systems, 547-550 (1997)＊　　　　　　　　　　　　　　　
1997

Munetomo, M., Takai, Y. and Sato, Y. : "StGA : An Application of a Genetic Algorithm to Stochastic Learning Automata", Systems and Computers in Japan, 27(10) : 68-78 (1997)＊　　　　　　　　　　　　　　　
1997

StGA : An Application of a Genetic Algorithm to Stochastic Learning Automata　　　　　　　　　　　　　　　
Masaharu Munetomo, Yoshiaki Takai, Yoshiharu Sato
Systems and Computers in Japan, 27, 10, 68, 78, Oct. 1996, [Peer-reviewed]
English, Scientific journal

Masaharu Munetomo, Yoshiaki Takai, and Yoshiharu Sato: "Genetic-Based Dynamic Load Balancing: Implementation and Evaluation", Parallel Problem Solving from Nature, Lecture Notes in Computer Science 1141, 920--929 (1996)*　　　　　　　　　　　　　　　
1996

Masaharu Munetomo, Yoshiaki Takai, and Yoshiharu Sato: "StGA: An application of a Genetic Algorithm to Stochastic Learning Automata", Systems and Computers in Japan, Vol.27, No.10, 68-78. (1996)*　　　　　　　　　　　　　　　
1996

「確率学習における遺伝的アルゴリズムの適用」　　　　　　　　　　　　　　　
『電子情報通信学会論文誌』, J79-D-II, 2, 230, 238, 1996

On tracking-ability of a stochastic genetic algorithm to changing environments
M Munetomo, Y Takai, Y Sato
INFORMATION INTELLIGENCE AND SYSTEMS, VOLS 1-4, 522, 526, 1996, [Peer-reviewed]
English, International conference proceedings

Genetic-based dynamic load balancing: Implementation and evaluation
Masaharu Munetomo, Yoshiaki Takai, Yoshiharu Sato
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 1141, 920, 929, Springer Verlag, 1996, [Peer-reviewed]
English, International conference proceedings

A Dynamic Ｌoad Balancing Scheme Using a Genetic Algorithm with Stochastic Learning
MUNETOMO Masaharu, TAKAI Yoshiaki, SATO Yoshiharu
IPSJ Journal, 36, 4, 868, 878, Information Processing Society of Japan (IPSJ), 15 Apr. 1995
Japanese, It is necessary to balance the load of each processor in a distributed system in order to utilize the system effectively. In a dynamic load balancing algorithm with distributed control, each processor observes load status of the system and dispatches tasks independently. We propose a dynamic load balancing scheme with distributed control which employs stochastic multicast messages. We encode a sending set of the requests for task dispatch into a binary string to which genetic operations with stochastic learning are applied in order to increase the probability for the requests to be accepted. Through simulation studies, we compared our scheme with some conventional load balancing methods. The results show the effectiveness of our scheme concerning mean response time of the tasks, success rate of the requests, and adaptability to environmental changes.

A Dynamic Load Balancing Scheme Using a Genetic Algorithm with Stochastic Learning
Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu
IPSJ Journal, 36, 4, 868, 878, Information Processing Society of Japan (IPSJ), 15 Apr. 1995
Japanese, It is necessary to balance the load of each processor in a distributed system in order to utilize the system effectively. In a dynamic load balancing algorithm with distributed control, each processor observes load status of the system and dispatches tasks independently. We propose a dynamic load balancing scheme with distributed control which employs stochastic multicast messages. We encode a sending set of the requests for task dispatch into a binary string to which genetic operations with stochastic learning are applied in order to increase the probability for the requests to be accepted...

A stochastic genetic algorithm for dynamic load balancing in distributed systems
M MUNETOMO, Y TAKAI, Y SATO
1995 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN AND CYBERNETICS, VOLS 1-5, 4, 3795, 3799, 1995, [Peer-reviewed]
English, International conference proceedings

An Efficient String Exchange Algorithm for a Subpopulation - Based Asynchronously Parallel Genetic Algorithm and Its Evaluation
MUNETOMO MASAHARU, TAKAI YOSHIAKI, SATO YOSHIHARU
IPSJ Journal, 35, 9, 1815, 1827, Information Processing Society of Japan (IPSJ), 15 Sep. 1994
Japanese, We Present an efficient string exchange scheme on subpopulaiton-based parallel genetic algorithms. The subpopulation-based parallel genetic algorithm divides a population into subpopulations in which genetic operations are executed simultaneously. In this scheme, exchanging strings between subpopulations through communicating network is essential to avoiding performance degradation of genetic search due to uniformity of the subpopulation. To reduce unnecessary inter-processor communications is an important issue to realize efficient parallel computation. In conventional subpopulation-based parallel genetic algorithms, however, inefficient parallel computations are taken place because string exchanges are executed at a constant interval or fully at random. The sigma-exchange algorithm we propose observes a fitness distribution of each subpopulation and starts on exchanging strings only when the standard deviation of the fitness distribution of some subpopulation decreases to some ratio. The purpose of our algorithm is to obtain more precise solutions with less inter-processor communications. We show the effectiveness of our scheme through simulation experiments on a multicomputer network in which communications are realized via asynchronous message passing.

A Genetic Approach to Dynamic Load Balancing in a Distributed Computing System
MUNETOMO M.
Proceedings of the First IEEE Conference on Evolutionary Computation, 1, 418, 421, Jun. 1994, [Peer-reviewed]
English, International conference proceedings

AN EFFICIENT MIGRATION SCHEME FOR SUBPOPULATION-BASED ASYNCHRONOUSLY PARALLEL GENETIC ALGORITHMS
M MUNETOMO, Y TAKAI, Y SATO
PROCEEDINGS OF THE FIFTH INTERNATIONAL CONFERENCE ON GENETIC ALGORITHMS, 649, 649, 1993, [Peer-reviewed]
English, International conference proceedings

Generating Collective Behavior of a Multi-Legged Robotic Swarm with Neuroevolution : A Case Study in a Two-landmark Navigation Problem
塚本遙日, 森本大智, 平賀元彰, 大倉和博, 棟朝雅晴, システム制御情報学会研究発表講演会講演論文集(CD-ROM), 65, 851, 858, 26 May 2021
システム制御情報学会, Japanese

「情報処理学会論文誌：数理モデル化と応用」の編集にあたって
棟朝 雅晴, 情報処理学会論文誌数理モデル化と応用（TOM）, 14, 1, i, i, 27 Jan. 2021
Japanese

Introducing modular networks based on multi-objective neuroevolution in rogue-like game
高橋寿徳, 棟朝雅晴, 情報処理学会研究報告(Web), 2021, MPS-132, 2021

Design of a model for estimating the difficulty of holds in sport climbing
西野直登, 桂大地, 棟朝雅晴, 棟朝雅晴, 情報処理学会研究報告(Web), 2021, MPS-132, 2021

「情報処理学会論文誌：数理モデル化と応用」の編集にあたって
棟朝 雅晴, 情報処理学会論文誌数理モデル化と応用（TOM）, 13, 2, i, i, 28 Aug. 2020
Japanese

Distribution System for Japanese Synthetic Population Data with Protection Level
Murata Tadahiko, Ichikawa Manabu, Goto Yusuke, Sugiki Akiyoshi, Date Susumu, Hanawa Toshihiro, Harada Takuya, Munetomo Masaharu, Hao Lee, Proceedings of the Fuzzy System Symposium, 36, 0, 269, 272, 2020

In this paper, we introduce a distribution system of synthesized data of Japanese population using Interdisciplinary Large-scale Information Infrastructures in Japan. Synthesized population is synthesized based on the statistics of census that are publicly released. Therefore, the synthesized data have no privacy data. However, since it is easy to estimate the compositions of households, working status in a certain area from the synthesized population, we distribute the synthesized data only for public or academic purposes. In the academic purposes, it is important to encourage young scholars to use a large-scale data of households, we define security levels for the attributes in the synthesized populations. According to the security levels, we distribute the data with proper attributes to applicants. We encourage researchers to use the synthetic populations to be familiar to large-scale data processing.

, Japan Society for Fuzzy Theory and Intelligent Informatics, Japanese

Generating Collective Behavior of Multi-legged Robotic Swarm with Deep Neuroevolution
MORIMOTO Daichi, HIRAGA Motoaki, OHKURA Kazuhiro, MATSUMURA Yoshiyuki, MUNETOMO Masaharu, Proceedings of the Annual Conference of JSAI, 2020, 0, 2M5OS3b02, 2M5OS3b02, 2020

In this paper, the controller of the multi-legged robotic swarm is designed by deep neuroevolution, which is a technique to train a deep neural network by using artificial evolution. The computer simulations are conducted with a 3D physics engine called Bullet. An aggregation task is examined with varying the sensor range to discuss the behavior. The results show that deep neuroevolution was able to generate collective behavior of the multi-legged robotic swarm. Moreover, the robotic swarm showed a potential behavior that might be useful to achieve more complex tasks.

, The Japanese Society for Artificial Intelligence, Japanese

「情報処理学会論文誌：数理モデル化と応用」の編集にあたって
棟朝 雅晴, 情報処理学会論文誌数理モデル化と応用（TOM）, 12, 3, i, i, 23 Dec. 2019
Japanese

北海道大学ハイパフォーマンスインタークラウドの設計,構築,運用まで (特集 総合力で造る情報基盤)
杉木 章義, 棟朝 雅晴, オペレーションズ・リサーチ = Communications of the Operations Research Society of Japan : 経営の科学, 64, 9, 507, 513, Sep. 2019
日本オペレーションズ・リサーチ学会 ; 1956-, Japanese

エッジコンピューティング環境における広域分散アプリケーションの多目的最適資源割当
藤田駿一, 棟朝雅晴, 情報処理学会研究報告(Web), 2019, CSEC-84, Vol.2019‐CSEC‐84,No.13,1‐6 (WEB ONLY), 25 Feb. 2019
Japanese

Introducing Modular Neural Networks in Learning Action Games
TAKAHASHI Toshinori, MUNETOMO Masaharu, Proceedings of the Annual Conference of JSAI, 2019, 0, 4Rin140, 4Rin140, 2019

In this paper, we introduce a modular network approach in neuroevolution for action game learning. We employ NEAT (NeuroEvolution of Augmenting Topologies) to generate modular networks learning game stages under different conditions, which are combined to obtain networks that can adapt to difficult situations appeared in actual game playing. We show the effectiveness of our approach in a Pygame instance compared to the original NEAT.

, The Japanese Society for Artificial Intelligence, Japanese

クラウドとエッジサーバを用いた広域分散アプリケーションの最適資源割り当てに関する検討
藤田駿一, 棟朝雅晴, 情報科学技術フォーラム講演論文集, 17th, 103‐106, 12 Sep. 2018
Japanese

実数値多目的最適化問題におけるリンケージ同定手法の性能検証
泉谷光祐, 棟朝雅晴, 電気学会電子・情報・システム部門大会講演論文集(CD-ROM), 2018, ROMBUNNO.MC1‐5, 05 Sep. 2018
Japanese

LSTMを用いたユーザの嗜好を反映するBGM自動作曲システム
岡部太亮, 棟朝雅晴, 情報処理学会研究報告(Web), 2018, MPS-118, Vol.2018‐MPS‐118,No.29,1‐2(WEB ONLY), 06 Jun. 2018
Japanese

隠消現実感アプリケーションに基づくエッジコンピューティングの性能推定
畑徹, 棟朝雅晴, 情報処理学会研究報告(Web), 2018, MPS-117, Vol.2018‐MPS‐117,No.23,1‐2 (WEB ONLY), 22 Feb. 2018
Japanese

Cgroupsを利用したHadoopにおける落ちこぼれタスクのリソース制限による再現
岩井良成, 杉木章義, 棟朝雅晴, 情報処理学会研究報告(Web), 2017, OS-140, Vol.2017‐OS‐140,No.13,1‐6 (WEB ONLY), 09 May 2017
Japanese

冬道情報サービス構築のためのエッジサーバを用いた分散処理フレームワークの検討
市居遼平, 棟朝雅晴, 情報処理学会研究報告(Web), 2017, ITS-68, Vol.2017‐ITS‐68,No.10,1‐7 (WEB ONLY), 21 Feb. 2017
Japanese

Webシステムの性能評価に基づくクラウド資源割当最適化モデルの提案
齋藤篤志, 山下雅喜, 三浦克宜, 棟朝雅晴, 情報処理学会研究報告(Web), 2017, MPS-112, Vol.2017‐MPS‐112,No.13,1‐6 (WEB ONLY), 20 Feb. 2017
Japanese

Management of Academic Cloud System at Hokkaido University
棟朝雅晴, 電子情報通信学会技術研究報告, 116, 124(ICM2016 8-23), 85‐86, 30 Jun. 2016
Japanese

クラウドブローカーシステムにおけるWeb三層システムの最適化数理モデルの検討
齋藤篤志, 三浦克宜, 棟朝雅晴, 電子情報通信学会技術研究報告, 116, 120(NC2016 6-15), 55‐56, 27 Jun. 2016
Japanese

Building an edge-based process framework for Winter Road Information Service
市居遼平, 棟朝雅晴, 杉木章義, 電子情報通信学会技術研究報告, 115, 504, 83, 88, 10 Mar. 2016
電子情報通信学会, Japanese

インタークラウド環境下での設計探査支援フレームワークにおける多目的設計最適化の適用
阿部友哉, 棟朝雅晴, 情報処理学会研究報告(Web), 2016, MPS-107, VOL.2016-MPS-107,NO.21 (WEB ONLY), 01 Mar. 2016
Japanese

自動車車載カメラ映像解析による降雪環境下での車両検出及び雪道路面推定手法の提案
岩坪潤, 棟朝雅晴, 情報処理学会研究報告(Web), 2016, MPS-107, VOL.2016-MPS-107,NO.22 (WEB ONLY), 01 Mar. 2016
Japanese

設計者の要求に基づく非劣解分析支援システムの提案
中野 翔, 渡邉 真也, 千葉 一永, 金崎 雅博, 棟朝 雅晴, 計測自動制御学会システム・情報部門学術講演会講演論文集, 2015, 631, 635, 18 Nov. 2015
多目的最適化によって得られた非劣解集合に対して設計者側の要求に基づいた傾向分析を実現する新たな分析支援システムの提案を行う．本システムは，従来までの相関ルールに基づく分析アプローチを改良したものであり，設計者側からの要求に即したルール抽出に特化することで，設計者の知りたい傾向をピンポイントで抽出することが可能となっている．, 計測自動制御学会, Japanese

述語論理式による仕様記述に基づくクラウドブローキングの提案
三浦克宜, 齋藤篤志, 棟朝雅晴, 情報処理学会研究報告(Web), 2015, MPS-105, VOL.2015-MPS-105,NO.6 (WEB ONLY), 22 Sep. 2015
Japanese

Toward Realizing a High-Performance Inter-Cloud System
棟朝雅晴, 電子情報通信学会技術研究報告, 115, 140, 43, 48, 16 Jul. 2015
電子情報通信学会, Japanese

クラウドブローカーのための抽象的なシステム記述の検討
三浦克宜, 齋藤篤志, 玉家武博, 棟朝雅晴, 電子情報通信学会技術研究報告, 115, 112(IBISML2015 1-26), 215, 220, 16 Jun. 2015
Japanese

インタークラウド環境における仮想システム構築の最適化サービスに関する検討
玉家 武博, 齋藤 篤志, 三浦 克宜, 棟朝 雅晴, 第77回全国大会講演論文集, 2015, 1, 177, 178, 17 Mar. 2015
クラウドサービスの多様化が進む中で、それぞれのユーザーが自身の要件に適したサービスを選択することが困難になりつつある。そこで本研究では、ユーザーの要件に適した、仮想マシン等からなる仮想システムを自動的に最適化し、構築するクラウドスケジューラの実現に向けた検討を行う。本研究で提案するシステムは，(1) インタークラウド環境におけるそれぞれのクラウドサービスに関する情報を提供するサービス，(2) ユーザーが構築する仮想システムの要求要件に基づいて最適化を行うサービス、(3) インタークラウド環境に仮想システムを自動的に構築するサービス、の3つのWebサービスによって構成される。, Japanese

Hadoopを用いた遺伝的アルゴリズムによるDDoS攻撃防止システムの検討
水越 大貴, 棟朝 雅晴, 情報処理学会研究報告. MPS, 数理モデル化と問題解決研究報告, 2014, 1, 1, 6, 02 Dec. 2014
インターネットにおけるセキュリティ上の脅威の一つとして DDoS 攻撃が深刻な問題になっている.DDoS 攻撃は一般に攻撃元の情報が改竄されているため,攻撃元の特定が非常に困難であり防ぐ事が難しい攻撃だといえる.また,攻撃パターンの学習によるパターンマッチングや,異常トラフィック検知などの手法が研究されているが,DDoS 攻撃において攻撃者はボットネット等を使用し、常に異なるトラフィックパターンの攻撃を仕掛けてくるため、過去のデータの解析から DDoS 攻撃を防ぐ事は非常に難しい.このような背景から,ネットワーク管理者は常に現在どのような攻撃が行われているかを監視、解析し,DDoS 攻撃に対処する必要性に迫られる.しかし,近年ネットワーク上を流れるトラフィック量は急激に増加しており,トラフィックの解析にはかなりの時間かかってしまう事が予測される.そこで本稿では,DDoS 攻撃に対し迅速に対処するシステムを作る事を目的とし,並列分散処理基盤である Hadoop 上での遺伝的アルゴリズムを使用したトラフィックパターンの解析手法を提案する。, 一般社団法人情報処理学会, Japanese

インタークラウド環境下での大規模分散設計最適化のための連携システムの設計
阿部友哉, 棟朝雅晴, 情報処理学会研究報告. MPS, 数理モデル化と問題解決研究報告, 2014, 18, 1, 2, 18 Sep. 2014
プライベートクラウドやパブリッククラウドを連携させたインタークラウド環境が整備され,仮想的に計算資源を無限に利用できるような環境が実現されつつある.本研究ではそのようなインタークラウド環境を想定して大規模かつ複雑な設計問題を扱う最適化フレームワークを構築する.具体的にはシミュレーションを実行するスパコン,大規模なパラメータサーベイを行うための最適化エンジンや分散データベース,解を視覚的に評価するための可視化装置などの複数のシステムの連携によって,設計問題に関する設計パラメータを統一的に管理,共有し最適化を行う.このフレームワークを構築するにあたって連携システムの詳細設計を行った., 一般社団法人情報処理学会, Japanese

クラウドPaaS上での多人数インタラクティブ遺伝的アルゴリズムによるスケーラブルな3Dモデリング
瀬山貴仁, 坂東信太郎, 棟朝雅晴, 情報処理学会研究報告. MPS, 数理モデル化と問題解決研究報告, 2014, 7, 1, 4, 14 Jul. 2014
本論文においては、ユーザーの負担を軽減するために、多人数が協調して解の評価を行うインタラクティブ進化計算の実装について議論する。実装にあたっては、クラウド PaaS(Platform as a Service) 基盤を前提としたスケーラブルなシステムを実現するためのシステム設計を行った。, 一般社団法人情報処理学会, Japanese

インタークラウド環境における大規模分散設計最適化フレームワークに関する検討
棟朝雅晴, 情報処理学会研究報告. BIO, バイオ情報学, 2014, 28, 1, 2, 18 Jun. 2014
プライベートクラウドおよびパブリッククラウドを全国規模で連携させたインタークラウド環境を想定し,大規模かつ複雑な設計問題の設計パラメータに関する情報を,設計者や最適化エンジンが共有,活用しつつ協調して設計を行うフレームワークについて検討する.具体的には,設計パラメータやその評価値等に関する情報を,スケーラブルな分散データベースシステム上に統合管理するとともに,シミュレーションプログラムや可視化システム等の連携を行うフレームワークを,インタークラウド環境における物理・仮想マシン群およびオブジェクトストレージを用いて実現するものである., 一般社団法人情報処理学会, Japanese

インタークラウド環境における大規模分散設計最適化フレームワークに関する検討(機械学習によるバイオデータマインニング,一般)
棟朝 雅晴, 電子情報通信学会技術研究報告. NC, ニューロコンピューティング, 114, 104, 165, 166, 18 Jun. 2014
プライベートクラウドおよびパブリッククラウドを全国規模で連携させたインタークラウド環境を想定し,大規模かつ複雑な設計問題の設計パラメータに関する情報を,設計者や最適化エンジンが共有,活用しつつ協調して設計を行うフレームワークについて検討する.具体的には,設計パラメータやその評価値等に関する情報を,スケーラブルな分散データベースシステム上に統合管理するとともに,シミュレーションプログラムや可視化システム等の連携を行うフレームワークを,インタークラウド環境における物理・仮想マシン群およびオブジェクトストレージを用いて実現するものである., 一般社団法人電子情報通信学会, Japanese

インタークラウド環境における大規模分散設計最適化フレームワークに関する検討
棟朝雅晴, 情報処理学会研究報告. MPS, 数理モデル化と問題解決研究報告, 2014, 28, 1, 2, 18 Jun. 2014
プライベートクラウドおよびパブリッククラウドを全国規模で連携させたインタークラウド環境を想定し,大規模かつ複雑な設計問題の設計パラメータに関する情報を,設計者や最適化エンジンが共有,活用しつつ協調して設計を行うフレームワークについて検討する.具体的には,設計パラメータやその評価値等に関する情報を,スケーラブルな分散データベースシステム上に統合管理するとともに,シミュレーションプログラムや可視化システム等の連携を行うフレームワークを,インタークラウド環境における物理・仮想マシン群およびオブジェクトストレージを用いて実現するものである., 一般社団法人情報処理学会, Japanese

HPCI先端ソフトウェア運用基盤の構築と運用
三浦信一, 滝澤真一朗, 松岡聡, 棟朝雅晴, 實本英之, 小林泰三, 情報処理学会研究報告. [ハイパフォーマンスコンピューティング], 2014, 30, 1, 6, 24 Feb. 2014
平成 24 年度より運用が開始されている HPCI では,スーパコンピュータ 「京」 や基盤センター群が保有するスーパコンピュータ間の認証基盤統一,データ共有を実現している.しかしながら,既存のスーパコンピュータシステムはバッチキューでジョブ管理されていることや,計算ノードでの管理者権限がないため,OS や分散システムの研究開発を行う CS 系ユーザの利用環境条件を満たさない.また,インターネット上より各種データを取得し,それを用いた計算を行う場合や,得られた成果を外部に公開するには,スーパコンピュータの利用は不向きである.そこで我々は,利用者に対してシステムへの管理者権限を付与する広域分散システムのホスティング機能を提供する,先端ソフトウェア運用基盤を HPCI の枠組みの中で構築し,平成 26 年 4 月より本格運用を開始する.本稿では先端ソフトウェア運用基盤の設計,構築及び運用について紹介する., 一般社団法人情報処理学会, Japanese

Multi-objective resource allocation optimization of WEB systems considering SLAs in distributed cloud environment
Takafumi Kawakatsu, Masaharu Munetomo, IPSJ SIG technical reports, 2013, 9, 1, 6, 04 Dec. 2013
We propose a mathematical model for optimal resource allocation of the WEB systems in distributed cloud environment. In the proposed model, we allocate WEB systems with load balancers to cloud systems that satisfy SLAs according to requests from users, which also have a capability of scale-out in increasing the overheads. We solve multi-objective optimization problems considering cost, response time, and throughput using a multi-objective genetic algorithm, and show the effectiveness of the proposed framework through experiments., Information Processing Society of Japan (IPSJ), Japanese

Multi-objective resource allocation optimization of WEB systems considering SLAs in distributed cloud environment
Takafumi Kawakatsu, Masaharu Munetomo, IPSJ SIG Notes, 2013, 9, 1, 6, 04 Dec. 2013
We propose a mathematical model for optimal resource allocation of the WEB systems in distributed cloud environment. In the proposed model, we allocate WEB systems with load balancers to cloud systems that satisfy SLAs according to requests from users, which also have a capability of scale-out in increasing the overheads. We solve multi-objective optimization problems considering cost, response time, and throughput using a multi-objective genetic algorithm, and show the effectiveness of the proposed framework through experiments., Information Processing Society of Japan (IPSJ), Japanese

Implementation of Multiple Classifier System on MapReduce Framework for Intrusion Detection
Masataka Mizukoshi, Shintaro Bando, Martin Schlueter, Masaharu Munetomo, IPSJ SIG Notes, 2013, 11, 1, 4, 15 Jul. 2013
Since the data volume from various facilities keeps growing rapidly in recent years, "big data" processing frameworks such as Hadoop have been developed as a scalable architecture to process large amount of data in cloud computing environment. We focus on intrusion detection problems which require large amount of data to be processed in order to detect malicious attacks. In this paper we discuss a Hadoop implementation of a multiple classifier system to enhance performances of the learning process in intrusion detection., Information Processing Society of Japan (IPSJ), English

広域分散ストレージ検証環境におけるI/O性能評価 (インターネットアーキテクチャ)
柏崎 礼生, 近堂 徹, 北口 善明, 楠田 友彦, 大沼 善朗, 中川 郁夫, 市川 臭平, 棟朝 雅晴, 高井 昌彰, 阿部 俊二, 横山 重俊, 下條 真司, 電子情報通信学会技術研究報告 : 信学技報, 112, 489, 105, 110, 14 Mar. 2013
大規模災害による危機意識の高まりから災害回復(Disaster Recover:DR)を実現するための技術として遠隔地データセンターでのバックアップや分散ストレージに注目が集まっている.現在我々はランダムアクセス性能の高さに特徴のある広域分散ストレージ環境を金沢大学,広島大学,Nilを中心として構築しており,本研究では本環境のI/O性能を評価し,この環境の有用性を示す., 一般社団法人電子情報通信学会, Japanese

広域分散ストレージ検証環境におけるI/O性能評価 (技術と社会・倫理)
柏崎 礼生, 近堂 徹, 北口 善明, 楠田 友彦, 大沼 善朗, 中川 郁夫, 市川 昊平, 棟朝 雅晴, 高井 昌彰, 阿部 俊二, 横山 重俊, 下條 真司, 電子情報通信学会技術研究報告 : 信学技報, 112, 488, 105, 110, 14 Mar. 2013
大規模災害による危機意識の高まりから災害回復(Disaster Recover:DR)を実現するための技術として遠隔地データセンターでのバックアップや分散ストレージに注目が集まっている.現在我々はランダムアクセス性能の高さに特徴のある広域分散ストレージ環境を金沢大学,広島大学,NIIを中心として構築しており,本研究では本環境のI/O性能を評価し,この環境の有用性を示す., 一般社団法人電子情報通信学会, Japanese

広域分散ストレージ検証環境におけるI/O性能評価
柏崎 礼生, 近堂 徹, 北口 善明, 楠田 友彦, 大沼 善朗, 中川 郁夫, 市川 昊平, 棟朝 雅晴, 高井 昌彰, 阿部 俊二, 横山 重俊, 下條 真司, 研究報告インターネットと運用技術（IOT）, 2013, 19, 1, 6, 07 Mar. 2013
大規模災害による危機意識の高まりから災害回復（Disaster Recover: DR）を実現するための技術として遠隔地データセンターでのバックアップや分散ストレージに注目が集まっている．現在我々はランダムアクセス性能の高さに特徴のある広域分散ストレージ環境を金沢大学，広島大学，NIIを中心として構築しており，本研究では本環境のI/O性能を評価し，この環境の有用性を示す．, Japanese

Hadoop環境上で動作する研究分野判定ツールの試作
平島慶典, 三浦克宜, 棟朝雅晴, 全国大会講演論文集, 2013, 1, 563, 565, 06 Mar. 2013
本研究では、与えられた論文に対して、的確な研究分野を判定するためのツールを開発する。研究を発展させる上で、関連研究のサーベイは重要であり、そのためには論文の適切な研究分野を知ることは極めて大切である。適切な研究分野を発見する方法として過去の論文と照らし合わせる方法が考えられる。しかしそれには膨大な計算量が掛かるため、逐次処理ではコストがかかる。この問題を解決するために、並列計算を使用している。研究分野の位置づけを行うために、本論文ではMahoutによるクラスタリングを行っており、そのための計算は、Hadoopを利用した並列計算を使用している。, 一般社団法人情報処理学会, Japanese

単峰性正規分布交叉を用いた実数値遺伝的アルゴリズムによる宇宙探査機の多重重力支援軌道最適化
田中一真, 棟朝雅晴, 赤間清, 全国大会講演論文集, 2013, 1, 469, 471, 06 Mar. 2013
複数天体の重力支援を利用した宇宙探査機の軌道最適化は,厳密解の発見が困難な,制約付き非線形多変数関数の最適化問題である.本研究では,欧州宇宙機関で公開されている軌道最適化問題を単峰性正規分布交叉を用いた実数値遺伝的アルゴリズムによって解く., 一般社団法人情報処理学会, Japanese

Design of Authentication System for High Performance Distributed Computing Environment
合田憲人, 東田学, 坂根栄作, 天野浩文, 小林克志, 棟朝雅晴, 江川隆輔, 建部修見, 鴨志田良和, 滝澤真一朗, 永井亨, 岩下武史, 石川裕, 情報処理学会論文誌トランザクション(CD-ROM), 2012, 2, 2013

進化型スワームロボットシステムにおける大規模並列計算環境を用いた非同期処理に関する一考察
竹中 貴治, 保田 俊行, 大倉 和博, 松村 嘉之, 棟朝 雅晴, 精密工学会学術講演会講演論文集, 2013, 0, 775, 776, 2013
スワームロボットシステムとは多数のロボットが局所的な環境・状況下での相互作用を通して群行動を創発するシステムである．本稿ではニューラルネットをCMA-ESにより最適化するCMA-NeuroESを制御器の設計に運用する．その際，進化の過程で膨大な計算コストが必要となるため，SMP Cluster型の大規模並列計算機で並列・高速化を図る．その上で，並列化を行う上で発生する同期待ち時間に着目し，実行時間短縮について考察する．, 公益社団法人 精密工学会, Japanese

クラウドコンピューティングを俯瞰する
梶田 将司, 棟朝 雅晴, 電子情報通信学会 通信ソサイエティマガジン, 7, 3, 166, 174, 2013
一般社団法人 電子情報通信学会, Japanese

Design of Authentication System for High Performance Distributed Computing Environment
合田 憲人, 東田 学, 坂根 栄作, 天野 浩文, 小林 克志, 棟朝 雅晴, 江川 隆輔, 建部修見, 鴨志田 良和, 滝澤 真一朗, 永井 亨, 岩下 武史, 石川 裕, 情報処理学会論文誌コンピューティングシステム（ACS）, 5, 5, 90, 102, 15 Oct. 2012
This paper presents design of the authentication system for the High Performance Computing Infrastructure (HPCI), which is currently deployed by the Ministry of Education, Culture, Sports, Science and Technology. The presented authentication system enables single sign-on to computers and shared storages on HPCI by utilizing the authentication mechanism on the Grid, "Grid Security Infrastructure (GSI)", and the identity federation mechanism, "Shibboleth". This paper also presents the experiments conducted on the testbed for the presented authentication system., 情報処理学会, Japanese

Multi-purpose optimization of resource allocation combined CDN and Web three-layer model in inter-cloud environment
KAWAKATSU TAKAFUMI, MUNETOMO MASAHARU, 情報処理学会研究報告(CD−ROM), 2012, 3, ROMBUNNO.MPS-90,NO.13, 15 Oct. 2012
Japanese

Development of Multiple Precision Arithmetic Software
YOSHIHARA IKUO, HONDA SHIORI, SAKAMOTO AI, YAMAMORI KUNIHITO, MUNETOMO MASAHARU, Mem Fac Eng Univ Miyazaki, 41, 337-341, 30 Jul. 2012
Japanese

Multiple-precision Calculation of .PI.
YOSHIHARA IKUO, SAKAMOTO AI, HONDA SHIORI, YAMAMORI KUNIHITO, MUNETOMO MASAHARU, Mem Fac Eng Univ Miyazaki, 41, 331-335, 30 Jul. 2012
Japanese

クラウドとビッグデータの活用がもたらすイノベーション (特集 ビジネスの変革を牽引するクラウドソリューション)
棟朝 雅晴, 堀田 多加志, 田中 誠司, 日立評論, 94, 7, 489, 491, Jul. 2012
日立評論社, Japanese

Report of a study for efficient multi-parameter survey of OCTA/cognac using ASNARO-RCM
萩田 克美, 棟朝 雅晴, 上島 豊, 計算工学講演会論文集 Proceedings of the Conference on Computational Engineering and Science, 17, 4p, May 2012
日本計算工学会, Japanese

446 Operation Planning of an Energy-Independent-House by a Plant-Shoot Solar Cell Module
Kawae Osamu, Obara Shinya, Munetomo Masaharu, The Proceedings of the Symposium on Environmental Engineering, 2012, 0, 393, 395, 2012
In this paper we consider the system to cover all energy by solar cells. System is composed of solar cells, PEFC, water electrolyzer, and heat pump. In this paper, we developed a analysis program by genetic algorithm. And, revealed the capacity of the equipment and two types of operational methods. A result of the analysis, equipment capacity needed in the first method of operating, solar cells is 250m2, PEFC is 8.1kW, water electrolyzer is 22.1kW, heat pump is 14.4kW, amount of hydrogen storage is 106.6kWh, amount of heat storage is 10.2kWh. In addition, equipment capacity needed in the second method of operating, solar cells is 300m2, PEFC is 4.7kW, water electrolyzer is 30.1kW, heat pump is 8.5kW, amount of hydrogen storage is 84.9kWh, amount of heat storage is 21.5kWh. To consider the cost of equipment as the next step., The Japan Society of Mechanical Engineers, Japanese

Multiple-precision Calculation of π
Yoshihara Ikuo, Sakamoto Ai, Honda Shiori, Yamamori Kunihito, Munetomo Masaharu, Memoirs of the Faculty of Engineering, University of Miyazaki, 41, 41, 331, 335, 2012
Multiple-precision calculation is necessary for precisely solving scientific engineering problems. Extremely long precision is employed to evaluate the mathematical constant, e.g. π, γ(Euler's constant), e(Nepier's constant) etc. To develope multiple-precision computing software, we try to calculate π with more than one million decimal digits. The proto-type code is verified by performing calculation with numerical examples and evaluated rapidness of calculation. Hother to π with 16,777,199 decimal digit is obtained., Miyazaki University, Japanese

Development of Multiple Precision Arithmetic Software
Yoshihara Ikuo, Honda Shiori, Sakamoto Ai, Yamamori Kunihito, Munetomo Masaharu, Memoirs of the Faculty of Engineering, University of Miyazaki, 41, 41, 337, 341, 2012
It is necessary to employ "multiple precision arithmetic" for computing long digit numbers, because numerical representation of computers is usually limited. This paper aims at making prototype software to compute more than one million digit numbers. A long digit number is divided into 2^n short digit numbers, each of which can be calculated by ordinal double precision arithmetic units. The key technique of "multiple precision arithmetic" is based on fast Fourier transform and convolution theorem. The prototype program is verified from the viewpoint of correctness of calculation up to 4 X 10^6 digits and speed up ratio vs theoretical value. The program is applied to calculation of a 10^7 or more digit π., Miyazaki University, Japanese

The effect of implementation of mixtures of Bayesian Network to dynamic environment problem
HORI SHIN'YA, MUNETOMO MASAHARU, AKAMA KIYOSHI, 情報処理学会研究報告(CD−ROM), 2011, 3, ROMBUNNO.MPS-85,NO.16, 15 Oct. 2011
Japanese

The effect of implementation of mixtures of Bayesian Network to dynamic environment problem
Shinya Hori, Masaharu Munetomo, Kiyoshi Akama, IPSJ SIG Notes, 2011, 16, 1, 6, 08 Sep. 2011
Bayesian Optimization Algorithm (BOA) can solve wide-spectrum of optimization problems by modeling their probabilistic models with conditional probabilities based on Bayesian networks. BOA succeeds in solving a variety of difficult optimization problems, however, it has not been applied successfully to dynamic environment because it tend to converge at one specific network and cannot adapt to change of probabilistic distributions. In this paper, we propose a BOA that introduces mixture distributions and inheritance of probability distribution of previous generation to adapt to dynamic envir..., Information Processing Society of Japan (IPSJ), Japanese

A GPU accelerated Fragment-Based De Novo Ligand Design by a Bayesian Optimization Algorithm
Mohamed Wahib, Asim Munawar, Masaharu Munetomo, Kiyoshi Akama, 研究報告数理モデル化と問題解決（MPS）, 2011, 6, 1, 6, 08 Sep. 2011
De Novo ligand design is an automatic fragment-based design of molecules within a protein binding site of a known structure. A Bayesian Optimization Algorithm (BOA), a meta-heuristic algorithm, is introduced to join predocked fragments with a user-supplied list of fragments. A novel feature proposed is the simultaneous optimization of force field energy and a term enforcing 3D-overlap to known binding mode(s). The performance of algorithm is tested on Liver X receptors (LXRs) using a library of about 14,000 fragments and the binding mode of a known heterocyclic phenyl acetic acid to bias the design. We further introduce the use of GPU (Graphical Processing Unit) to overcome the excessive time required in evaluating each possible fragment combination. We show how the GPU utilization enables experimenting larger fragment sets and target receptors for more complex instances. The Results show how the nVidia's Tesla C2050 GPU was utilized to enable the generation of complex agonists effectively. In fact, eight of the 1809 molecules designed for LXRs are found in the ZINC database of commercially available compounds.De Novo ligand design is an automatic fragment-based design of molecules within a protein binding site of a known structure. A Bayesian Optimization Algorithm (BOA), a meta-heuristic algorithm, is introduced to join predocked fragments with a user-supplied list of fragments. A novel feature proposed is the simultaneous optimization of force field energy and a term enforcing 3D-overlap to known binding mode(s). The performance of algorithm is tested on Liver X receptors (LXRs) using a library of about 14,000 fragments and the binding mode of a known heterocyclic phenyl acetic acid to bias the design. We further introduce the use of GPU (Graphical Processing Unit) to overcome the excessive time required in evaluating each possible fragment combination. We show how the GPU utilization enables experimenting larger fragment sets and target receptors for more complex instances. The Results show how the nVidia's Tesla C2050 GPU was utilized to enable the generation of complex agonists effectively. In fact, eight of the 1809 molecules designed for LXRs are found in the ZINC database of commercially available compounds., Information Processing Society of Japan (IPSJ), English

RL-004 Deployment of Contents Management Systems on Hokkaido University Academic Cloud
Munetomo Masaharu, Takai Yoshiaki, 情報科学技術フォーラム講演論文集, 10, 4, 15, 18, 07 Sep. 2011
Forum on Information Technology, Japanese

Design of Advanced Software Deployment Infrastructure in HPCI Wide-area Distributed Environment
TAKIZAWA SHIN'ICHIRO, MUNETOMO MASAHARU, UNO ATSUYA, KOBAYASHI TAIZO, JITSUMOTO HIDEYUKI, MATSUOKA SATOSHI, MATSUOKA SATOSHI, ISHIKAWA YUTAKA, 情報処理学会研究報告(CD−ROM), 2011, 2, ROMBUNNO.HPC-130,NO.68, 15 Aug. 2011
Japanese

Design of Advanced Software Deployment Infrastructure in HPCI Wide-area Distributed Environment
Shinichiro Takizawa, Masaharu Munetomo, Atsuya Uno, Taizo Kobayashi, Hideyuki Jitsumoto, Satoshi Matsuoka, Yutaka Ishikawa, IPSJ SIG Notes, 2011, 68, 1, 7, 20 Jul. 2011
The purpose of HPCI, which will be operated from autumn 2012, is to support HPC researchers to use K supercomputer, and its initial services are a federated authentication and global file sharing between K and supercomputers provided by computer centers in Japan. However, supercomputers are not suitable for HPC system researchers as their operations do not give users enough privileges. We design the advanced software deployment infrastructure that hosts distributed systems where researchers can have administrator privileges. We introduce the design of the system and a precedent system imple..., Information Processing Society of Japan (IPSJ), Japanese

A Light Framework for the Unified Representation and Execution of Variant Tasks in a Grid Based Environment
WAHIB MOHAMED, MUNAWAR ASIM, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 79, 9, I1, I11, 12 Jul. 2010
Grid computing has gained a wide interest from the research community over the past one and a half decade. The immense effort has resulted in mature tools and technologies for grid computing. The utilization of experience and tools of grid computing in the next generation of distributed systems (e.g. cloud com-puting) is a logical step. However, many problems that come along with grid computing do limit such an effort. Among these problems is the sophistication of each production grid to a specific task type, size and dependency. In other words, grid computing in practice up to the moment c..., Information Processing Society of Japan (IPSJ), English

An effective Bayesian Network Construction Method for BOA
HORI SHIN'YA, MUNETOMO MASAHARU, AKAMA KIYOSHI, 情報処理学会研究報告(CD−ROM), 2009, 6, ROMBUNNO.MPS-77,23, 15 Apr. 2010
Japanese

An effective Bayesian Network Construction Method for BOA
堀 伸哉, 棟朝 雅晴, 赤間 清, 情報処理学会研究報告, 2009, 6, 1, 7, Apr. 2010
本論文は BOA の改良型アルゴリズム、EBOA （Effective BOA） を提案する。EBOA は BOA のベイジアンネットワーク構築フェイズにおいて、探索する遺伝子ノード数をエントロピーの値によってクラスタリングし、それぞれのクラスターでベイジアンネットワークを構築することで遺伝子の探索時間を減少させる。また、これと同時にエントロピーの値による探索遺伝子数の絞込みも導入する。これら二つの手法を取り入れた EBOA は BOA において問題となる多大な実行時間を短縮することで複雑で巨大な問題の最適化を行うことが可能となる。This paper proposes an Effective Bayesian Optimization Algorithm (EBOA) which improves network construction process of BOA. EBOA performs clustering of locus nodes based on their entropy measures and at the same time, removes loci that are not necessary to be modeled based on their entropy. After the clustering, it constructs a Bayesian Network in each cluster to reduce the running time for searching networks for BOA to solve large and complex optimization problems., 情報処理学会, Japanese

An effective Bayesian Network Construction Method for BOA
Shinya Hori, Masaharu Munetomo, Kiyoshi Akama, IPSJ SIG Notes, 2010, 23, 1, 7, 25 Feb. 2010
This paper proposes an Effective Bayesian Optimization Algorithm (EBOA) which improves network construction process of BOA. EBOA performs clustering of locus nodes based on their entropy measures and at the same time, removes loci that are not necessary to be modeled based on their entropy. After the clustering, it constructs a Bayesian Network in each cluster to reduce the running time for searching networks for BOA to solve large and complex optimization problems., Information Processing Society of Japan (IPSJ), Japanese

A Grid based Unified Framework for Optimization
MUNAWAR ASIM, WAHIB MOHAMED, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 69, 41, 23, 26, 09 May 2008
This paper presents a Service Oriented Architecture (SOA) compliant Problem Solving Environment (PSE) that allows the user to implement any metaheuristics based algorithm over a Grid. We call this framework a Grid based Unified Framework for Optimization (GridUFO). GridUFO provides a unified approach for sharing and using metaheuristics algorithms (solvers) and objective functions over a Grid in an easiest possible "plug & play" manner. In this way the user can take all the advantages of the Grid without taking into consideration any of the complexities posed by the Grid environment. The framework is accessible to the user through a Web service or through a fully integrated 2nd generation Web portal. GridUFO provides well-defined interfaces between the user-programmed objective functions and solvers. We also present MetaHeuristics Markup Language (MHML), an XML based language that acts as an interface between the user and the framework. In this paper we will discuss the design and implementation of GridUFO in detail. Moreover, we will talk about our experience of working with Grids, and we will also make some recommendations for future research., Information Processing Society of Japan (IPSJ), English

An Empirical Study on BOA which Introduces Scatter Search
SATAKE YUTA, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 68, 17, 109, 112, 04 Mar. 2008
Bayesian Optimization Algorithm (BOA) builds its probabilistic model which represents distribution of promising solutions and generates new solutions based on the model. Because BOA detects interdependent loci by using the model, it can solve a wide spectrum of optimization problems effectively. BOA which introduces local search in order to enhance its performance is already proposed. However, they did not use Scatter Search, which can create new search points effectively, as local search. In this paper, we propose BOA which introduces scatter search and discuss its effectiveness., Information Processing Society of Japan (IPSJ), Japanese

Parallelization of a genetic algorithm using linkage identification and context dependent crossover
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG technical reports, 2007, 128, 171, 174, 20 Dec. 2007
Parallelized competent genetic algorithms (cGAs) which can detect problem structures automatically can give us problem solving environments for a wide spectrum of real-world problems. As such algorithms, the DBOA, PBOA and parallel BOA which are based on BOA and pLINC which is based on LINC had been proposed. However, model buildings in the parallelized BOAs are performed under some restrictions or using backtracking to obtain feasible models. While pLINC can be parallelized in a simpler way, the original LINC spends huge number of fitness evaluations. In this paper, we try to parallelize D^5 which can also be parallelized in a simple way and requires relatively small number of fitness evaluations. We also try to parallelize population evolution with context dependent crossover (CDC) which can combine overlapping building blocks effectively. Moreover, we perform some experiments to compare the paralleled D^5-GA+CDC and other cGAs to reveal their properties., Information Processing Society of Japan (IPSJ), Japanese

Parallelization of a genetic algorithm using linkage identification and context dependent crossover
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 67, 128, 171, 174, 20 Dec. 2007
Parallelized competent genetic algorithms (cGAs) which can detect problem structures automatically can give us problem solving environments for a wide spectrum of real-world problems. As such algorithms, the DBOA, PBOA and parallel BOA which are based on BOA and pLINC which is based on LINC had been proposed. However, model buildings in the parallelized BOAs are performed under some restrictions or using backtracking to obtain feasible models. While pLINC can be parallelized in a simpler way, the original LINC spends huge number of fitness evaluations. In this paper, we try to parallelize D^5 which can also be parallelized in a simple way and requires relatively small number of fitness evaluations. We also try to parallelize population evolution with context dependent crossover (CDC) which can combine overlapping building blocks effectively. Moreover, we perform some experiments to compare the paralleled D^5-GA+CDC and other cGAs to reveal their properties., Information Processing Society of Japan (IPSJ), Japanese

A Proposal of Crossover Method for Complex Building Blocks Overlapping
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI, 情報処理学会論文誌数理モデル化と応用（TOM）, 48, 15, 23, 33, 15 Oct. 2007
In order to realize effective genetic algorithms, there have been several techniques to identify linkage sets of loci to form a building block (BB) (Heckendorn, et al.). By contrast, the way to realize effective crossover from the linkage information given by such techniques has not been studied enough. Especially for problems with overlapping BBs, a crossover method proposed by Yu, et al. (2005) is the first and only known research. However it cannot perform well for problems with complexly overlapping BBs due to BB disruptions and insufficient variety of crossover sites. In this paper, we propose a crossover method which examines values of given parental strings to determine which variables are exchanged to produce new and different strings without increasing BB disruptions as much as possible. Because the proposed method considers the context of parental strings, it is called context dependent crossover (CDC). Combining a scalable linkage identification technique and the CDC, an effective algorithm for problems with overlapping BBs is provided. Moreover, to test the proposed method, we design test functions with controllable complexity of overlaps., Information Processing Society of Japan (IPSJ), Japanese

Consideration of Estimation of Distribution Algorithms employing Simulated Annealing
MAEDA HARUKI, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 64, 43, 13, 16, 17 May 2007
Estimation of Distribution Algorithms(EDAs) build a probabiblistic model which reflects promising solutions and generate new solutions using the model. Considering the dependencies between the variables with the model, they can slove GA-difficult problems. On the other hand, EDAs have a drawback they need extensive computational cost for building the model. While the development of such optimization methods is advanced, the hybridizations of these methods and local search methods are widely studied. In this paper, we propose a hybrid algorithm of Bayesian Optimization Algorithm(BOA), which is one of EDAs, and a local search employing Simulated Annealing(SA)., Information Processing Society of Japan (IPSJ), Japanese

Asim Munawar, Masaharu Munetomo, Akama Kiyoshi　　　　　　　　　　　　　　　
1191, 1198, 2007

Considering ECGA which Introduces Tabu Search
SATAKE YUTA, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG technical reports, 2006, 135, 61, 64, 21 Dec. 2006
Extended Compact Genetic Algorithm (ECGA) builds its probabilistic model which represents distribution of promising solutions and generates new solutions based on the model. Because ECGA detects interdependent loci by using the model, it can solve a wide spectrum of optimization problems effectively. ECGA which introduces neighbourhood search in order to enhance its performance is already proposed. However, they did not use tabu search, which is one of the most effective neighbourhood searches, as neighbourhood search. In this paper, we propose ECGA which introduces tabu search and consider its effectiveness., Information Processing Society of Japan (IPSJ), Japanese

Considering ECGA which Introduces Tabu Search
SATAKE YUTA, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 62, 135, 61, 64, 21 Dec. 2006
Extended Compact Genetic Algorithm (ECGA) builds its probabilistic model which represents distribution of promising solutions and generates new solutions based on the model. Because ECGA detects interdependent loci by using the model, it can solve a wide spectrum of optimization problems effectively. ECGA which introduces neighbourhood search in order to enhance its performance is already proposed. However, they did not use tabu search, which is one of the most effective neighbourhood searches, as neighbourhood search. In this paper, we propose ECGA which introduces tabu search and consider its effectiveness., Information Processing Society of Japan (IPSJ), Japanese

Tabu Searchを導入したECGAについての検討
佐竹 佑太, 棟朝 雅晴, 赤間 清, 情報処理学会研究報告, 2006, 135, 61, 64, 21 Dec. 2006
情報処理学会, Japanese

Considering Computational Cost of Probabilistic Model-Building Genetic Algorithms which Introduce Local Search
SATAKE YUTA, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 59, 56, 25, 28, 25 May 2006
Probabilistic Model-Building Genetic Algorithms (PMBGAs) build their probabilistic models which represent distribution of promising solutions and generate new solutions based on the models. Although PMBGAs can solve a wide spectrum of optimization problems effectively, their probabilistic model-buiding processes need extensive computational cost. On the other hand, in PMBGAs which introduce local search, although the number of evaluations is high, they can reduce the cost to build models. In this paper, we consider the relation of the number of evalations and the cost to build models in PMBGAs which introduce local search., Information Processing Society of Japan (IPSJ), Japanese

Generation of Correct Parallel Programs for Solving Constraint Satisfaction Problems
SAITO YUSUKE, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 2006, 20, 103, 108, 27 Feb. 2006
Based on Equivalent Transformation (ET) computation model, a theory of parallel program generation is proposed, where correct parallel programs are generated from a set of ET-rules that transform a set of definite clauses that represents a given problem. Correctness of the method is explained. Usefulness of the method is partly shown by experiments of a parallel program generated for a constraint solving problem., Information Processing Society of Japan (IPSJ), Japanese

Generation of Correct Parallel Programs for Solving Constraint Satisfaction Problems
SAITO YUSUKE, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 167, 20, 103, 108, 27 Feb. 2006
Based on Equivalent Transformation (ET) computation model, a theory of parallel program generation is proposed, where correct parallel programs are generated from a set of ET-rules that transform a set of definite clauses that represents a given problem. Correctness of the method is explained. Usefulness of the method is partly shown by experiments of a parallel program generated for a constraint solving problem., Information Processing Society of Japan (IPSJ), Japanese

Applying the D^5-GA for problems with overlapping building blocks
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 2005, 93, 65, 68, 21 Sep. 2005
The D^5-GA [4] detects interdepended loci and use the information as a guidance on exploration. The existing D^5-GA divides all loci into exclusive sets-called linkage sets-of loci which are linked tightly to construct a building block and exchanges loci in a same set simultaneously to enhance efficient building block mixing. However, some real-world problems have overlapping building blocks. In this paper, we extend the structure of the linkage sets to apply the D^5-GA for problems with overlapping building blocks. In addition, we modify a crossover method proposed by Yu et al. [6] to achieve (1) smaller BB disruptions and (2) larger BB exchanges., Information Processing Society of Japan (IPSJ), Japanese

Linkage Identification for Real-Coded Genetic Algorithms based on Additive Decomposability and Difference Signature Independency of Objective Function
TEZUKA MASARU, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 53, 20, 9, 12, 09 Mar. 2005
In the case that a problem is decomposable to a number of sub-problems which can be optimized independently, the problem is solved effectively by optimizing sub-problems separately. In optimization problems by means of genetic algorithms, a set of loci of which each sub-problem consists is called linkage group. Linkage identification is the method which recognizes linkage groups. In this paper, we define the linkage of Real-Coded GAs clearly. Then we propose two new linkage identification methods, LINC-R and LIDI-R, directly based on the definition. LINC-R is based on additive decomposability of a objective function and LIDI-R is based on the independency of the signature of difference. These methods effectively identify linkages. Parallel optimization of the linkage groups has a capability to obtain better solutions in smaller number of function evaluations than conventional GAs., Information Processing Society of Japan (IPSJ), Japanese

Structure Prediction of Protein by BOA with parallel network construction
MURAO NAOYA, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 162, 19, 67, 72, 07 Mar. 2005
Bayesian optimization algorithm (BOA) is an advanced optimization algorithm that effectively solves GA-difficult problems in which it is difficult to ensure tight encoding. BOA constructs Bayesian networks based on probabilistic distributions of current promising solutions, and generates next population based on the obtained networks. However, computational cost of the estimation depends on the problem size, and is known the calculation cost increases rapidly along increasing the problem size. In our previous study, we proposed a BOA with parallelized Bayesian network construction. In this paper, we try to solve proteins structure prediction problems that minimize the structual energy by employing BOA with the parallel network construction, and investigate the effectiveness of our approach., Information Processing Society of Japan (IPSJ), Japanese

リンケージ同定GAを導入した波長割当方式の評価(ネットワークプロトコル)
釘本 健司, 棟朝 雅晴, 赤間 清, 情報処理学会研究報告. EIP, [電子化知的財産・社会基盤], 2004, 89, 63, 69, 02 Sep. 2004
次世代のインターネットの基盤ネットワークとして,WDM(Wavelength Division Multiplexing)に基づいたフォトニックネットワーク(WDM-PN)が注目されている.このWDM-PNにおいては,物理トポロジ上に光パスを割り当てることで論理トポロジが構成される.限られた波長を効率良く使って論理トポロジを構成する問題は,波長割当問題と呼ばれ,制約つき組合わせ問題の一つである.本稿では,リンケージ同定を導入した遺伝的アルゴリズムの波長割当問題への適用を試みたので報告する.本アルゴリズムでは,トラフィック全体の遅延の最小化を目的とし,リンケージ同定手法としてLINC(Linkage Identification by Nonlinearity Check)およびLIEM(Linkage Identification with Epistasis Measure)を用いた.計算機シミュレーションにより単純遺伝的アルゴリズムとの比較を行ない,LIEMの効果を確認した., 一般社団法人情報処理学会, Japanese

An Evaluation of WDM Wavelength Assignment using Genetic Algorithm with Linkage Identification
Kugimoto Takeshi, Munetomo Masaharu, Akama Kiyoshi, IPSJ SIG Notes, 2004, 89, 63, 69, 02 Sep. 2004
Wavelength Division Multiplexing (WDM) is a technology which multiplexes the light of different wavelength to single optical fiber. It allows more efficient use of an optical firer, but also allows new network feature - virtual topology. Virtual topology is the connection graph whose links are lightpaths. In such the WDM network, virtual network topology can be dynamically constructed by changing the combination of the wavelength assignment in physical topology. The wavelength assignment problem is a combinational optimization problem. Heuristic method such as genetic algorithm (GA) is often used to solve such an optimization problem. In this paper, we consider the Linkage Identification based GA approach. Linkage Identification is a sort of technique to detect Building Block in genes. The results obtained from Linkage Identification based GA approach are compared with ordinary simple GA to illustrate the effectiveness of proposed approach., Information Processing Society of Japan (IPSJ), Japanese

A-024 Real-Coded Genetic Algorithm to Optimize Fitness Function with Sampling Error
Tezuka Masaru, Munetomo Masaharu, Akama Kiyoshi, 情報科学技術フォーラム一般講演論文集, 3, 1, 55, 56, 20 Aug. 2004
Forum on Information Technology, Japanese

Empirical Investigations on the parallelized Bayesian Optimization Algorithm
MURAO NAOYA, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 157, 20, 169, 174, 01 Mar. 2004
The Bayesian optimization algorithm is an advanced search technique generating population of string by estimating distribution of promising solutions, which can solve difficult problems efficiently that simple Genetic algorithm cannot solve. However, since computational cost for estimating distribution is large, research for parallelizing of BOA is necessary. Although Parallel BOA is experimented in the previous research, it is performed on a small number of processors. In this paper, empirical investigation on the performance of parallel BOA is performed in detail., Information Processing Society of Japan (IPSJ), Japanese

Performance comparisons between parallel linkage identification and parallel BOA
MURAO NAOYA, MUNETOMO MASAHARU, AKAMA KIYOSHI, 情報処理学会研究報告. MPS, 数理モデル化と問題解決研究報告, 47, 0, 57, 60, 11 Dec. 2003
Japanese

Modeling dependency from distributions of strings classified according to fitness change
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI, 情報処理学会研究報告. MPS, 数理モデル化と問題解決研究報告, 47, 0, 61, 64, 11 Dec. 2003
Japanese

Performance comparisons between parallel linkage identification and parallel BOA
MURAO NAOYA, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 47, 122, 57, 60, 11 Dec. 2003
The linkage identification and the Bayesian Optimization Algorithm are proposed to solve GA-difficult problems. These algorithms are competent, while their computational cost is high. Recently, parallelized version of linkage identification and BOA are studied to decrease time to obtain solutions. However, it is no clear which algorithms should we use to solve the problem. In this paper, we compare the performance of parallel linkage identification to that of parallel BOA, and then we consider a relation between the nature of problem and their performances to obtain a guideline to select parallel GAs., Information Processing Society of Japan (IPSJ), Japanese

Modeling dependency from distributions of strings classified according to fitness change
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 47, 122, 61, 64, 11 Dec. 2003
Genetic Algorithms (GAs) can perform effectively by identifying a set of loci tightly linked and strongly interdepended to form a same building block. Various methods are alreadly proposed to detect such dependencies. Some of them investigate the fitness changes from the perturbation of gene value and some others estimate the distribution of strings in promising sub population. In this paper, we propose a new method combining both of them, which detects dependencies through the estimation of the distribution of strings classified according to fitness change. The proposed method can detect dependencies accurately by a little more than O(l) fitness evaluations., Information Processing Society of Japan (IPSJ), Japanese

Investigations on Parallelizing Methods of Evolutionary Computation
MURAO NAOYA, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 93, 29, 161, 166, 11 Mar. 2003
Parallel genetic algorithm (PGA) is an active research area in evolutionary computation and a number of papers have been published, however, relation between the nature of the problems and the method to be applied to solve them. Most papers deal with a tailered version of PGA to solve a specific problem. In this paper, we perform empirical investigations on the relation between characteristics of problems based on building blocks weights and selection of PGA methods. We also compare conventional methods with a novel approach of PGA based on linkage identification. Final goal of the paper is to show a guideline to users choosing one of the PGA methods based on the characteristics of the problems to be solved., Information Processing Society of Japan (IPSJ), Japanese

MAN Design by Hierarchical Linkage Genetic Algorithm
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 110, 108, 73, 78, 21 Nov. 2002
Metropolitan area network design is a difficult combinatorial optimization problem, which generates an optimal network topology that satisfies geographical constraints, user traffic constraints, etc. from a huge number of candidates. In the network design problems, simple GAs sometimes fail to recombine building blocks, which leads to unappropriate solutions. In the previous work [7], a GA based on linkage identification could solve this problem effectively, which identifies a set of loci belonging to a same linkage group to exchange building blocks accurately. In this paper, single layer linkage identification has been extended to hierarchical one, which considers interdependence of building blocks in a recursive manner, instead of identifying interdependence of loci., Information Processing Society of Japan (IPSJ), Japanese

A Genetic Algorithm using Linkage Identification for Metropolitan Area Network Design
TSUJI MIWAKO, MUNETOMO MASAHARU, AKAMA KIYOSHI, IPSJ SIG Notes, 39, 36, 9, 12, 10 May 2002
In genetic algorithms, it is important to encode strings ensuring tight linkage for their huilding blocks. in network design prohlems, however, it is difficult to encode strings appropriately because network design is dependent not only on geographical constraints hut also on other complex factors such as bias on traffic demands, routing policy, and so on. Although there's many applications of genetic algorithms to network topology design, most of them haven't paid attention to tight encoding of building blocks, or considered only geographical characteristics. In order to realize tight linkage among loci and realize effective genetic search, this paper introduces LIEM (Linkage Identification with Epistasis Measure) - a technique for identifying linkage sets, sets of loci tightly liked to form building filocks - to realize effective network design. Through empirical studies, we show the effectiveness of the network design with the LIEM compared to that with conventional genetic algorithms., Information Processing Society of Japan (IPSJ), Japanese

Designing a Distributed Algorithm Using Genetic Algorithms for Bandwidth Allocation -Recovery Mechanism from Node Failure -
小林 英博, 棟朝 雅晴, 赤間 清, 佐藤 義治, マルチメディア通信と分散処理ワークショップ論文集, 2000, 15, 91, 96, 06 Dec. 2000
Japanese

Designing a Distributed Algorithm Using Genetic Algorithms for Bandwidth Allocation : Recovery Mechanism from Link Failure
KOBAYASHI Hidehiro, MUNETOMO Masaharu, AKAMA Kiyoshi, SATO Yoshiharu, IPSJ SIG Notes, 100, 102, 7, 12, 26 May 2000
Network resources are limited in their capacity, moreover fast high capacity communication links are expensive. Thus bandwidth allocation and efficient usage of limited resources are important. Previously, GRA(Genetic Routing Algorithm)has been proposed as a bandwidth allocation with genetic algorithm. This algorithm is centralized algorithms and execute localizing optimizing, in case of trouble it is impossible to allocate bandwidth. Therefore we suggested distributed-GRA(D-GRA)previously, however this is imperfect about communication links failure and recovering from trouble. Then, in this paper, we eill improve D-GRA to recover from communication link failure., Information Processing Society of Japan (IPSJ), Japanese

Controling Frequency of Route Estimation in Routing Algorithm
YAMAGUCHI Naohiko, MUNETOMO Masaharu, AKAMA Kiyoshi, SATO Yoshiharu, IPSJ SIG Notes, 100, 102, 69, 74, 26 May 2000
Expansion of computer networks such as the Internet is so rapid and traffic over the networks is increasing along their expansion. Therefore, it is important to utilize network resources efficiently and reduce network latency by distributing communication packets among alternative routes. To realize effective load balancing, we have proposed a network routing algorithms employing genetic algorithms to generate alternative routes and observe loads of links along the routes adaptively. In this paper, we will focus on evaluation strategy of routes. Changing evaluation strategy and frequency of observations, we examine their effects on overall network load status., Information Processing Society of Japan (IPSJ), Japanese

動的負荷分散機構を有する分散型経路制御アルゴリズム
山口 直彦, 棟朝 雅晴, 佐藤 義治, 第60回全国大会講演論文集, 2000, 1, 449, 450, 14 Mar. 2000
Japanese

遺伝的アルゴリズム4<分担 : 北野 宏明 編>　　　　　　　　　　　　　　　
産業図書, 2000

Bandwidth Allocation using Genetic Algorithms
KOABYASHI Hidehiro, MUNETOMO Masaharu, SATO Yoshiharu, IPSJ SIG Notes, 95, 94, 91, 96, 27 May 1999
Effective bandwidth allocation becomes essential to utilize network resources, espacially in large networks. In allocation problems, it is difficult to obtain optimal solution in reasonable computational cost because they are classified into combinatorial optimization problems. Mario Gerlra et. al[1] proposed an allocation algorithm that minimize the mean packet delay. In this paper, we try to solve a multi-objective optimization problem that minimizes the delay and also minimizes its variance. We employ genetic algorithms with multi-objective selection strategies in order to obtain a sef of Pareto optima., Information Processing Society of Japan (IPSJ), Japanese

Inter-AS Routing with Evolutionary methods
YAMAGUCHI Naohiko, MUNETOMO Masaharu, SATO Yoshiharu, IPSJ SIG Notes, 95, 94, 97, 102, 27 May 1999
In this paper, we propose an adaptive inter-AS routing algorithm that has a load balancing mechanism that distributes communication packets among alternative routes generated by genetic operators. The Border Gateway Protocol (BGP) widely employed for inter AS routing does not take an adaptive approach: it cannot determine routes based on current load status of the network although it distributes packets among alternative routes with the same distance measure. Our algorithm called GIAR (Genetic Inter AS Routing) observes load status of links along the alternative routes adaptively to realize load balancing among them by distributing packets probabilistically among them. To realize robust observation of routes in inter-AS routing, we introduce a threshold policy that classifies load status of links based on its queue length., Information Processing Society of Japan (IPSJ), Japanese

Identifying linkage groups by nonlinearity/non-monotonicity detection
MUNETOMO Masaharu, Proceedings of the 1999 Genetic and Evolutionary Computation Conference, 1999

A prediction algorithm for process resurce usage based on a state transition model
Moriguchi Shuichi, Munemoto Masaharu, Sato Yoshiharu, 全国大会講演論文集, 56, 0, 58, 59, 17 Mar. 1998
Information Processing Society of Japan (IPSJ), Japanese

A migration scheme for the genetic adaptive routing algorithm
M Munetomo, Y Takai, Y Sato, 1998 IEEE INTERNATIONAL CONFERENCE ON SYSTEMS, MAN, AND CYBERNETICS, VOLS 1-5, 2774, 2779, 1998
English

Identifying linkage by nonlinearity check
MUNETOMO Masaharu, Technical Report IlliGAL Report, 1998
University of Illinois

Designing a genetic algorithm using the linkage identification by nonlinearity check
MUNETOMO Masaharu, Technical Report IlliGAL Report, 1998
University of Illinois

A Meeting Scheduling System Supported by Mobile Agents
TOMIKAWA Yuki, TAKAI Yoshiaki, MUNETOMO Masaharu, YAMAMOTO Tsuyoshi, IPSJ SIG Notes, 85, 104, 133, 136, 06 Nov. 1997
When we negotiate an urgent issue through WAN such as the Internet, the unpredictable network latency may cause inefficient communication between negotiators. Because of the rapid traffic increase of the Internet, it is more and more difficult to communicate quickly. Mobile agents are programs that can autonomously travel across a network. Arriving at the remote computer, they can interact with other agents locally. Hence, the mobile agents are suitable for negotiation through WAN. In this paper we discuss a framework for making out a meeting schedule by using mobile agents. We implement a prototype system using Aglets, and we discuss agent's ability for more available communication system., Information Processing Society of Japan (IPSJ), Japanese

A Load Sharing Mechanism of Adaptive Routing Algorithms
MUNETOMO Masaharu, TAKAI Yoshiaki, SATO Yoshiharu, IPSJ SIG Notes, 80, 13, 205, 210, 30 Jan. 1997
This paper presents an adaptive routing algorithm which has a load balancing mechanism among alternative paths, and shows the effectiveness of the algorithm through simulation experiments. Conventional routing algorithms broadcast information on routing tables or link status in a network, which leads to consume much communication cost when the becomes large. A routing algorithm we propose generates alternative paths and perform evaluation of communication delay only for paths frequently used. This mechanism greatly reduces communication cost for information exchanging of the routing. In the algorithm, we employ genetic algorithms in generating alternative paths among which we distribute communication packets to ensure load balancing. We perform simulation experiments using a simulator of network communications and the result of the experiments says that an effective routing is achieved by less communication cost., Information Processing Society of Japan (IPSJ), Japanese

Adaptive Source Routing Using a Genetic Algorithm
MURAI Yasunori, MUNETOMO Masaharu, TAKAI Yoshiaki, IPSJ Technical Report, 79, 108, 43, 48, 14 Nov. 1996
We propose an adaptive algorithm for network routing and show the effectiveness of it through simulation experiments. The algorithm, which is based on a genetic algorithm, dynamically changes routes to keep the communication delay minimal by observing the communication delay of the packets. Every router on the network has the populations of genes which represent the candidates of routes. We use the mean communication delay along the routes to calculate genetic fitness of them. The result of the experiments shows that the algorithm is able to fit a large scale network and performance of it strongly depends on initial conditions of the genes., Information Processing Society of Japan (IPSJ), Japanese

Adaptive Routing using a Genetic Algorithm
Murai Yasunori, Munetomo Masaharu, Takai Yoshiaki, 全国大会講演論文集, 52, 0, 121, 122, 06 Mar. 1996
計算機ネットワークの拡大とトラヒックの増大に伴い、通信経路を決定するルーティング手法が急速にその重要性を高めている。本稿では遺伝的アルゴリズムを応用して、ネットワーク状態の変化に適応し、動的に経路選択を行うルーティング手法を提案する。経路選択の目的は平均の通信遅延時間を最小にすることであるが、これに要する付加的な制御情報の通信もネットワークのトラヒックに影響を与えるため、その通信は最小限に押さえられる必要がある。, Information Processing Society of Japan (IPSJ), Japanese

Evaluation of a Dynamic Load Balancing Algorithm by using Multicast on a Massively Parallel Processor
Ikeda Masaki, Munetomo Masaharu, Takai Yoshiaki, 全国大会講演論文集, 52, 0, 161, 162, 06 Mar. 1996
分散システムの利用率を向上させるためには、システムを構成する計算機間で負荷を平均化する必要がある。動的負荷分散アルゴリズムは負荷の重い計算機から負荷の軽い計算機へタスクを転送することでシステム全体として負荷の平均化をはかる。少ない通信量で効果的なタスク転送を行なうために、タスク転送要求の送出先を複数指定するマルチキャストを導入した手法が提案されている。 一方、局所メモリを持つ自立した計算ノードが専用の高速通信ネットワークにより相互結合されたMIMD型の並列計算機である超並列計算機は、高速な並列計算機を安価に実現するアーキテクチャとして近年注目を集めており、数多くの開発例が存在する。 そこで本論文では、マルチキャストによる動的負荷分散アルゴリズムを超並列計算機へ実装し、そのアルゴリズムの性能評価を行なうことで、超並列計算機上でのアルゴリズムの特性を調べる。具体的には、マルチキャストを用いた動的負荷分散アルゴリズムのシミュレータをParallel-Ware(ExPress)の通信ライブラリを用いて超並列計算機SR-2001上に実現し、シミュレーション実験を通してアルゴリズムの性能評価を行なう。, Information Processing Society of Japan (IPSJ), Japanese

An Application of a Genetic Algorithm to Stochastic Learning
MUNETOMO Masaharu, TAKAI Yoshiaki, SATO Yoshiharu, The Transactions of the Institute of Electronics,Information and Communication Engineers., 79, 2, 230, 238, 25 Feb. 1996
確率的な環境への適応学習を行う場合, 確率学習オートマトンに代表される強化学習が一般に用いられるが, 選択可能な行動の数が多くなった場合に最適解への収束が著しく遅くなるという欠点がある. 本論文では遺伝的アルゴリズムを応用することで, 強化学習における収束速度の問題点を解消する手法を提案する. 提案するアルゴリズムStGA(Stochastic Genetic Algorithm)においては, すべての可能な行動の中から少数の行動を集団としてサンプリングし, その集団に対して確率学習オートマトンを適用することで強化学習の収束速度を向上させる. 更に, 遺伝的操作を用いて集団内に含まれていない新たな行動を生成することを通して集団の内容を更新し, 最適な行動を効率良く探索する. StGAの収束性を証明するため, 確率学習オートマトンのε-optimalityをもとにした理論的解析を行う. 更にシミュレーション実験により, 可能な行動の数が多い場合におけるStGAの有効性を示す., The Institute of Electronics, Information and Communication Engineers, Japanese

Emergence of Cooperative Strategies in a Multi Agent Game
Tomikawa Yuki, Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, 情報処理学会研究報告. 人工知能研究会報告, 95, 105, 25, 30, 07 Nov. 1995
An SLA (Stochastic Learning Automaton) with huge number of possible Strategies takes a lot of time to converge. To avoid this problem ; an StGA (Stochastic Genetic Algorithm) was proposed to accelerate the learning process of the SLA. In this paper, we apply the StGA to strategy acquisition in games played by the groups of agents. The agents have implicit communication capability where there is no a pri ori meaning defined for the messages. We discuss the possibility of emergence of cooperation strategies through simulation experiments., Information Processing Society of Japan (IPSJ), Japanese

Dynamic Load Balancing with Genetic Algorithms on a UNIX Network
Yamashita Takayuki, Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, 全国大会講演論文集, 50, 0, 249, 250, 15 Mar. 1995
複数の計算機をLANで接続して資源の共有を図る分散システムにおいて、計算機間で負荷の分散を行うことにより、応答時間の短縮や資源利用率の改善など、システム性能の向上を図ることができる。この目的のため、種々の負荷分散方式が提案されてきた。負荷分散方式は、静的負荷分散方式と動的負荷分散方式に分類することができる。さらに、動的負荷分散は、負荷情報の管理とタスク転送の決定を一台の計算機で行う集中制御型と、各計算機で独立して行う分散制御型に分けられる。本研究では[3]を基に、マルチキャストによるタスク転送要求の送出先の決定に対して遺伝的アルゴリズムを適用することで、より効率的な負荷情報の収集と利用を図る分散制御型動的負荷分散方式を提案する。また、UNIXネットワークで構成される分散システム上に実装し、模擬タスクを用いたシミュレーション実験により性能評価を行う。, Information Processing Society of Japan (IPSJ), Japanese

A proof of convergence on genetic algorithms with elitist schemes using inhomogeneous Markov chains
Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, 全国大会講演論文集, 50, 0, 285, 286, 15 Mar. 1995
本稿ではエリート戦略を有する遺伝的アルゴリズム(Genetic Algorithms,GA)に関して、非斉次マルコフ連鎖を用いた解析を行う。GAの収束性に関しては、マルコフ連鎖を用いた解析が従来行われてきたが、本論文では、非斉次マルコフ連鎖の遷移行列を用い、より簡明な収束性の証明を行う。さらにその結果を用いて、大域的最適解を得る確率に関する収束速度の下限を求めた。, Information Processing Society of Japan (IPSJ), Japanese

An application of a stochastic genetic algorithm to strategy acquisition in games played by groups
Tomikawa Yuki, Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, 情報処理学会研究報告. 人工知能研究会報告, 95, 23, 85, 90, 06 Mar. 1995
In this paper we discuss strategy acquisition in games played by the groups of agents. In such games, we have a huge number of possible of possible strategies because each agent in a group could take a different strategy. We employ StGA(a Stochastic Genetic Algorithm) which evaluates fitness values by using a stochastic learning automaton in order to realize effective learning in stochastic environments. The StGA samples a small number of strategies from all possible ones and applies stochastic learning and genetic operations to the sampled strategies. Through simulation experiments, we show the effectiveness of the StGAin the strategy acquisition., Information Processing Society of Japan (IPSJ), Japanese

An Application of a Stochastic Genetic Algorithm to Strategy Acquisition in Games
Tomikawa Yuki, Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, Bulletin of the Faculty of Engineering,Hokkaido University, 172, p15, 22, Feb. 1995
Hokkaido University, Japanese

A genetic load balancing scheme based on an internal model of nodes in distributed systems
Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, IPSJ SIG Notes, 94, 105, 31, 36, 02 Dec. 1994
In this paper, we present a dynamic load balancing algorithm which learns a sending set of task migration requests. The algorithm is based on an internal model of nodes cousisting of FIFO and Round-Robin queues. In dynamic load balancing in general, we equalize each processor's load by migrating tasks from heavily-loaded processors to lightly-loaded ones. If we send task migration requests randomly or by a broadcast, many unnecessary messages will be sent. The proposed algorithm employs multicast messages which are sent to specified nodes. The sending set of the messages is coded into a genetic string to which genetic operations are applied., Information Processing Society of Japan (IPSJ), Japanese

A Genetic Algorithm which has a Fitness Evaluation Mechanism by Stochastic Learning(1) : BasicModel
Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, 全国大会講演論文集, 49, 0, 231, 232, 20 Sep. 1994
従来の遺伝的アルゴリズム(Genetic Algorithms,以下GAと略す)では、正確な適合度値が必要なときに必要な数だけ求められることを暗黙の前提としている。しかし、実際の問題へ応用する場合、適合度評価に時間を要し、一度に多くの適合度値を計算することが現実的でないことがある。また、確率的な環境への適応学習などの場合、環境から得られる情報は、ある行動の成功・失敗の2値で示されるため、適合度の値として直接採用することはできない。本論文では、逐次的に適合度評価を行なうことで、確率的な環境に適応する遺伝的アルゴリズムStGA(Stochastic Genetic Algorithm)を提案する。StGAでは適合度の評価に確率学習オートマトン(Stochastic Learning Automata,SLA)を採用した。これにより、環境から得られる情報が成功・失敗の2値に限られ、かつ逐次的にしか評価値が得られない場合でも、適切な適合度値の分布を集団内に作り出す。また、StGAをSLAの改良とみなすこともできる。SLAには、状態空間のサイズが非常に大きな場合に、収束が著しく遅くなるという欠点がある。この欠点を改善するために、従来、連想記憶を用いた状態空間の圧縮などの対策が講じられてきたが、問題に依存した静的な方法であることから一般に広く用いることはできない。StGAでは、状態空間をGAの個体の形にコーディングし、集団という形で状態空間からサンプリングを行なって、それに対して遺伝的操作を適用することにより、問題に適応する形で状態空間の圧縮を実現することができる。本論文では、SLAのみの場合とシミュレーション実験により比較検討することを通して、GAによる状態空間の圧縮が収束性の向上に大きな効果があることを示す。, Information Processing Society of Japan (IPSJ), Japanese

A Genetic Algorithm which has a Fitness Evaluation Mechanism by Stochastic Learning(2) : AnApplicatingtoStrategyAcquisitionofGames
Tomikawa Yuki, Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, 全国大会講演論文集, 49, 0, 233, 234, 20 Sep. 1994
従来の遺伝的アルゴリズム(Genetic Algorithm,以下GAと略す)を、適合度値の評価に時間を要する問題や確率環境への適応学習に適用することは困難である。このような問題に対して、確率学習による適合度評価機構を有する遺伝的アルゴリズムStGA(stochastic Genetic Algorithm)が提案されている。StGAでは、適合度の評価に確率学習オートマトンSLA(stochastic Learning Automata)を用いている。SLAには、状態空間のサイズが非常に大きい場合に収束が著しく遅くなるという欠点がある。StGAはこの点を改善し、問題に適応する形で状態空間を圧縮することを目的としている。我々はStGAが状態空間の圧縮を行なうという点に着目し、これを戦略の種類が非常に多いゲームにおける戦略の獲得に応用できるのではないかと考えた。本論文では、StGAとSLAをゲームにおける戦略の獲得を行なう手段としてインプリメントして対戦を行ない、状態空間のサイズが大きい場合におけるStGAの有効性の検証を行なう。, Information Processing Society of Japan (IPSJ), Japanese

A Dynamic Loaad Balancing System using Genetic Algorithms and Stochastic Learning Automata
Yamashita Takayuki, Munetomo Masaharu, Takai Yoshiaki, Sato Tishiharu, 全国大会講演論文集, 49, 0, 251, 252, 20 Sep. 1994
複数の計算機をLANで接続して資源の共有を図る分散システムにおいて、計算機間で負荷の分散を行うことにより、応答時間の短縮や資源利用率の改善など、システム性能の向上を図ることができる。この目的のため、種々の負荷分散方式が提案されてきた。分散制御型の動的負荷分散方式に遺伝的操作を導入した手法として、遺伝的アルゴリズムと確率学習オートマトンによる動的負荷分散(GeSLA)に関する研究が行われている。この手法においては、タスク転送をどの計算機に対して要求するかを記述した文字列を遺伝的アルゴリズム(Genetic Algorithms,GA)における個体とし、その適合度値の更新に確率学習オートマトン(Stochastic Learning Automata,SLA)による確率的山登り法を適用している。本研究では、UNIXワークステーションをLANで接続した分散システム上にこの方式を実装し、実際に生成したタスクを用いた実験により性能評価を行う。, Information Processing Society of Japan (IPSJ), Japanese

An Efficient String Exchange Algorithm for a Subpopulation-Based Asynchronously Parallel Genetic Algorithm and Its Evaluation
Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, IPSJ Journal, 35, 9, 1815, 1827, 15 Sep. 1994
We Present an efficient string exchange scheme on subpopulaiton-based parallel genetic algorithms. The subpopulation-based parallel genetic algorithm divides a population into subpopulations in which genetic operations are executed simultaneously. In this scheme, exchanging strings between subpopulations through communicating network is essential to avoiding performance degradation of genetic search due to uniformity of the subpopulation. To reduce unnecessary inter-processor communications is an important issue to realize efficient parallel computation. In conventional subpopulation-based ..., Information Processing Society of Japan (IPSJ), Japanese

A Dynamic Load Balancing Scheme Using Stochastic Learning Automata and Genetic Algorithms
Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, 情報処理学会研究報告. 人工知能研究会報告, 94, 20, 95, 102, 08 Mar. 1994
A distributed computing system is a collection of autonomous computers loosely connected via a communicating network whose latency is relatively large. In improving the system performance, it is important to keep the load of each processor even. On a distributed dynamic load balancing algorithm, each processor observes their load state and sends a task in order to balance their loads. In our scheme, we use a population of strings each of which stands for a set of processors to which requests of a task migration are sent, and genetic operations and stochastic learning are applied in order to realize efficient task migrations. Through empirical investigations using a simulator, we show the effectiveness of our scheme., Information Processing Society of Japan (IPSJ), Japanese

A Genetic Scheme for Distributed Dynamic Load Balancing
Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, Bulletin of the Faculty of Engineering,Hokkaido University, 167, 167, p127, 135, Jan. 1994
Hokkaido University, Japanese

A Cooperative Search Strategy Using Hierarchical Genetic Algorithms and Its Implementation on the UNIX-Network
Takahashi Masakazu, Munetomo Masaharu, Takai Yoshiaki, Sato Yoshiharu, 情報処理学会研究報告. 人工知能研究会報告, 93, 103, 9, 16, 24 Nov. 1993
In a subpopulation-based parallel genetic algorithm (PGA), a population is divided into subpopulations to which genetic operations are applied simultaneously. We propose a hierarchical subpopulation-based PGA model that has an optimization mechanism of GA parameters for each subpopulation using a meta-level GA. In our model, parameters for genetic operations applied to a subpopulation are coded into a string referred to as a meta-code which is optimized via meta-level GA operations. Through exchanging information concerning meta-codes and subpopulations, the model realizes an effective search on a distributed computing environment. We implement the model on a local area network that consists of UNIX workstations., Information Processing Society of Japan (IPSJ), Japanese

A Parallel Genetic Algorithm Based on the Adaptive Coding
TAKAHASHI Masakazu, MUNETOMO Masaharu, TAKAI Yosiaki, SATO Yoshiharu, 全国大会講演論文集, 46, 0, 301, 302, 01 Mar. 1993
遺伝的アルゴリズムは、生物の遺伝子の働きにヒントを得た最適化手法である。問題の対する多数の解候補を遺伝子の形にコーディングし、その適応度の高いものが増加してゆく「淘汰」(selection)2つの遺伝子内の部分情報を交換する「交叉」(crossover)、ある小さな確率で遺伝子内の情報が変化する「突然変異」(mutation)の基本3操作を一世代とし、それを繰り返すことによって近似最適解を得ようとするアルゴリズムである。しかし、一般的にコーディングやcrossover方法の設定に関しては、ビルディングブロック仮説を満たす必要がある。しかしながら。問題によってはこの仮説を常に満たすようなコーディング方法を求めることが困難な場合がある。その為、最適解に収束しない事も少なくない。そこで本稿では、コーディングを動的に変化させるadaptive codingを提案しナップザック問題を用いた数値実験によりその有効性を確認する。, Information Processing Society of Japan (IPSJ), Japanese

An Efficient Exchange Algorithm for Parallel GAs and Its Evaluation
棟朝 雅晴, 高井 昌彰, 佐藤義治, 情報処理学会研究報告アルゴリズム（AL）, 1992, 58, 41, 48, 17 Jul. 1992
遺伝子集団分割による並列遺伝的アルゴリズムにおける遺伝子交換アルゴリズムの改良とその評価を行なった結果について述べる。遺伝子集団の中に含まれるスキーマ数の減少を、その適合度の分布の標準偏差を計算することで間接的に評価し、遺伝子交換を行なう基準とすることで、遺伝子集団の中に含まれているスキーマの数の急速な減少を防ぎ、遺伝的アルゴリズムのスキーマ処理効率を維持する。そのために標準偏差と遺伝子集団内に含まれるスキーマの数について、Sigma?Hypothesisと呼ばれる仮説を立て、理論的解析とシミュレーション実験によりこれを検証した。さらに、この仮説に基づく遺伝子交換アルゴリズムについて、その有効性を実験により確認した。Improvement of exchange algorithms on parallel genetic algorithms and their evaluation are presented in this paper. To avoid rapid decrease of the number of schemata in a population and maintain efficient schema processing, the algorithm presented estimates indirectly the decrease of the number of schemata by using standard deviation of fitness distribution. We propose sigma-hypothesis and discuss its validity through some theoretical and empirical results. A parallel genetic algorithm based on the hypothesis is presented and some experiments are made to confirm its efficiency., Japanese

アカデミッククラウド調査報告書2012 (新産業調査レポートシリーズ)
吉岡 信和, 棟朝 雅晴, 本橋 賢二, 西村 一彦, 谷沢 智史, 横山 重俊
インプレスR&D, 24 Aug. 2012, 210, [Joint work]

Advances in Grid Computing　　　　　　　　　　　　　　　
Mohamed Wahib, Asim Munawar, Masaharu Munetomo, Kiyoshi Akama, pp.19-28
2011, [Contributor]

Genetic algorithms - theory and advanced methods
Masaharu Munetomo
森北出版, Jul. 2008, 4627847815, 160, [Single work]

Advances in Evolutionary Algorithms　　　　　　　　　　　　　　　
Mohamed Wahib, Asim Munawar, Masaharu Munetomo, Kiyoshi Akama, pp.315-334
IN-TECH, 2008, [Contributor]

Linkage in Evolutionary Computation　　　　　　　　　　　　　　　
Asim Munawar, Mohamed Wahib, Miwako Tsuji, Masaharu Munetomo, Kiyoshi Akama, pp.159-187, pp.441-459
Springer, 2008, [Contributor]

Computational Intelligence Paradigms – Innovative Applications　　　　　　　　　　　　　　　
Miwako Tsuji, Masaharu Munetomo, pp.251-280
Springer, 2008, [Contributor]

統計データ科学事典　　　　　　　　　　　　　　　
棟朝雅晴, 遺伝的アルゴリズム
2007, [Contributor]

Evolutionary Computation in Dynamic and Uncertain Environments　　　　　　　　　　　　　　　
Masaru Tezuka, Masaharu Munetomo, Kiyoshi Akama, pp.417-436
2007, [Contributor]

遺伝的アルゴリズム4　　　　　　　　　　　　　　　
棟朝雅晴, 第９章：遺伝的アルゴリズムによる適応ルーティング
産業図書, 2000, 4782851499, [Contributor]

Telecommunications Optimisation: Heuristic and Adaptive Methods　　　　　　　　　　　　　　　
Masaharu Munetomo, pp.151-166
John Weily & Sons, 2000, [Contributor]

Computational Intelligence in Telecommunication Networks　　　　　　　　　　　　　　　
Masaharu Munetomo, pp.287-302
CRC Press, 2000, [Contributor]

Optimal Feature Selection for Intrusion Detection Systems Employing Multi-Objective Genetic Algorithm　　　　　　　　　　　　　　　
Chang Geng, Masaharu Munetomo
2018 JPNSEC International Workshop on Evolutionary Computation, 31 Aug. 2018, English, Oral presentation

クラウドコンピューティングを用いた進化型ロボティックスワームの群れ行動生成
森川達矢, 保田俊行, 大倉和博, 松村嘉之, 棟朝雅晴
計測自動制御学会システムインテグレーション部門講演会(CD-ROM), 14 Dec. 2015, Japanese

インタークラウド環境下での大規模分散設計最適化のための最適化実行基盤の設計と実装
阿部友哉, 棟朝雅晴
計測自動制御学会システム・情報部門学術講演会講演論文集(CD-ROM), 18 Nov. 2015, Japanese

設計者の要求に基づく非劣解分析支援システムの提案
中野翔, 渡邉真也, 千葉一永, 金崎雅博, 棟朝雅晴
計測自動制御学会システム・情報部門学術講演会講演論文集(CD-ROM), 18 Nov. 2015, Japanese

インタークラウド環境における仮想システム構築の最適化サービスに関する検討
玉家武博, 斎藤篤志, 三浦克宜, 棟朝雅晴
情報処理学会全国大会講演論文集, 17 Mar. 2015, Japanese

有翼式宇宙往還機の安定性を考慮した亜音速飛行時の空力設計
倉田優太, 金崎雅博, 千葉一永, 渡邉慎也, 棟朝雅晴
数値流体力学シンポジウム講演論文集(CD-ROM), 2015, Japanese

分散クラウドシステムにおける遠隔連携技術　　　　　　　　　　　　　　　
棟朝 雅晴
学際大規模情報基盤共同利用・共同研究拠点 第1回ネットワーク型学際研究シンポジウム, 11 Mar. 2014
[Invited]

インターネット上のデータ利活用を促進するための人間ベース遺伝的アルゴリズム　　　　　　　　　　　　　　　
幸田里奈, 長谷部良輔, 大西圭, 棟朝雅晴
第6回進化計算学会研究会, 07 Mar. 2014

アカデミッククラウドにおけるCloudStackの活用事例と今後の展望　　　　　　　　　　　　　　　
棟朝 雅晴
CloudStack Day Japan 2014, 06 Mar. 2014
[Invited]

研究支援に係るアカデミッククラウドの調査検討　　　　　　　　　　　　　　　
棟朝 雅晴
平成25年度国家課題対応型研究開発推進事業『アカデミッククラウド環境構築に係るシステム研究』提案「コミュニティで紡ぐ次世代大学ICT環境としてのアカデミッククラウド」最終報告会, 13 Feb. 2014

研究支援に係るアカデミッククラウド　　　　　　　　　　　　　　　
棟朝 雅晴
大学ICT推進協議会年次大会「コミュニティで紡ぐ次世代大学ICT環境としてのアカデミッククラウド」事業中間報告, 18 Dec. 2013

研究支援のためのアカデミッククラウド　　　　　　　　　　　　　　　
棟朝 雅晴
アカデミッククラウドシンポジウム2013, 05 Sep. 2013

単峰性正規分布交叉を用いた実数値遺伝的アルゴリズムによる宇宙探査機の多重重力支援軌道最適化
田中一真, 棟朝雅晴, 赤間清
全国大会講演論文集, 06 Mar. 2013, Japanese
複数天体の重力支援を利用した宇宙探査機の軌道最適化は,厳密解の発見が困難な,制約付き非線形多変数関数の最適化問題である.本研究では,欧州宇宙機関で公開されている軌道最適化問題を単峰性正規分布交叉を用いた実数値遺伝的アルゴリズムによって解く.

Hadoop環境上で動作する研究分野判定ツールの試作
平島慶典, 三浦克宜, 棟朝雅晴
全国大会講演論文集, 06 Mar. 2013, Japanese
本研究では、与えられた論文に対して、的確な研究分野を判定するためのツールを開発する。研究を発展させる上で、関連研究のサーベイは重要であり、そのためには論文の適切な研究分野を知ることは極めて大切である。適切な研究分野を発見する方法として過去の論文と照らし合わせる方法が考えられる。しかしそれには膨大な計算量が掛かるため、逐次処理ではコストがかかる。この問題を解決するために、並列計算を使用している。研究分野の位置づけを行うために、本論文ではMahoutによるクラスタリングを行っており、そのための計算は、Hadoopを利用した並列計算を使用している。

クラウドコンピューティングの最新動向　　　　　　　　　　　　　　　
棟朝 雅晴
OR学会北海道支部講演会, 18 Feb. 2013
[Invited]

北海道大学アカデミッククラウドの活用事例　　　　　　　　　　　　　　　
棟朝 雅晴
学術情報基盤オープンフォーラム「大学クラウド活用における、検証と課題と対策」, 08 Feb. 2013
[Invited]

北海道大学アカデミッククラウドのご紹介とクラウド技術の最新動向，研究動向について　　　　　　　　　　　　　　　
棟朝 雅晴
第2回デバイスとクラウドの高度融合による新事業創出研究会, 24 Jan. 2013
[Invited]

リンケージツリー遺伝的アルゴリズムにおける計算量削減の検討　　　　　　　　　　　　　　　
鈴木一史, 棟朝雅晴
進化計算シンポジウム2012講演論文集, Dec. 2012

スワームロボットシステムにおける大規模並列計算環境を用いた分散型CMA-ESの実装　　　　　　　　　　　　　　　
竹中貴治, 保田俊行, 大倉和博, 松村嘉之, 棟朝雅晴
進化計算シンポジウム2012講演論文集, Dec. 2012

クラウド環境における進化計算用グリッドサービスの並列化効率の評価　　　　　　　　　　　　　　　
藤田二夫, 保田俊行, 大倉和博, 松村嘉之, 伍賀正典, 棟朝雅晴
進化計算シンポジウム2012講演論文集, Dec. 2012

異なるクラウド管理ソフトウェア環境の相互接続方式に関する検討
相澤孝至, 棟朝雅晴
情報処理北海道シンポジウム講演論文集, 06 Oct. 2012, Japanese

ASNARO‐RCMを用いたOCTA/cognacのパラメータサーベイの効率化に関する報告
萩田克美, 棟朝雅晴, 上島豊, 大宮学
計算工学講演会論文集(CD−ROM), 29 May 2012, Japanese

Hokkaido University Academic Cloud: Largest Academic Cloud System in Japan　　　　　　　　　　　　　　　
Masaharu Munetomo
Cloud Technical Leadership Forum, 10 May 2012, English, Invited oral presentation
[Invited], [International presentation]

Report of a study for efficient multi-parameter survey of OCTA/cognac using ASNARO-RCM
萩田 克美, 棟朝 雅晴, 上島 豊
計算工学講演会論文集 Proceedings of the Conference on Computational Engineering and Science, May 2012, Japanese

北海道大学アカデミッククラウドの構築とサービスについて　　　　　　　　　　　　　　　
棟朝雅晴
アカデミッククラウドワークショップ2012＠広島, 08 Feb. 2012, Japanese, Keynote oral presentation
[Invited], [Domestic Conference]

北海道大学における大規模学術クラウドの構築と運用について　　　　　　　　　　　　　　　
棟朝雅晴
サイエンティフィック研究会システム技術分科会第２回会合, 27 Jan. 2012, Japanese, Invited oral presentation
[Invited], [Domestic Conference]

北海道大学アカデミッククラウドの構築と運用について　　　　　　　　　　　　　　　
棟朝雅晴
グリッド協議会第33回ワークショップ, 21 Dec. 2011, Japanese, Keynote oral presentation
[Invited], [Domestic Conference]

北海道大学アカデミッククラウド〜国内最大規模の学術クラウドについて　　　　　　　　　　　　　　　
棟朝雅晴
Open Cloud Conference 2011 in Sapporo, 09 Dec. 2011, Japanese, Keynote oral presentation
[Invited], [Domestic Conference]

Estimation of Distribution Algorithms without Explicit Selections　　　　　　　　　　　　　　　
Masaharu Munetomo
The 8th World Multi-Conference on Systemics, Cybernetics and Informatics, 2004, English, Invited oral presentation
[Invited], [International presentation]

Linkage Identification by Non-monotonicity Detection for Overlapping Functions　　　　　　　　　　　　　　　
Masaharu Munetomo
Journal Showcase at the 2000 Genetic and Evolutionary Computation Conference, 2000, English, Invited oral presentation
[Invited], [International presentation]

Designing Genetic Algorithms for Adaptive Routing Algorithms in the Internet　　　　　　　　　　　　　　　
Masaharu Munetomo
Workshop on Evolutionary telecommunications: Past, present, and future, at the 1999 Genetic and Evolutionary Computation Conference, 1999, English, Invited oral presentation
[Invited], [International presentation]

一般教育演習　　　　　　　　　　　　　　　
北海道大学

知識系工学特論　　　　　　　　　　　　　　　
北海道大学

情報処理II　　　　　　　　　　　　　　　
北海道大学

情報処理I　　　　　　　　　　　　　　　
北海道大学

情報科学　　　　　　　　　　　　　　　
北海道大学

知能情報学特論　　　　　　　　　　　　　　　
室蘭工業大学

情報数理科学特別講義　　　　　　　　　　　　　　　
大阪府立大学

科学技術英語演習　　　　　　　　　　　　　　　
北海道大学

情報工学実験第一　　　　　　　　　　　　　　　
北海道大学

情報学II　　　　　　　　　　　　　　　
北海道大学

計算システム設計学特論　　　　　　　　　　　　　　　
北海道大学

ソフトウェア方法論　　　　　　　　　　　　　　　
北海道大学

メディアコンテンツ工学　　　　　　　　　　　　　　　
北海道大学

ネットワークとクラウド　　　　　　　　　　　　　　　
北海道大学

基礎数学　　　　　　　　　　　　　　　
北海道工業大学

情報システム設計学特論　　　　　　　　　　　　　　　
北海道大学

A novel development of evolutionary hybrid hyperheuristics
Grants-in-Aid for Scientific Research
01 Apr. 2024 - 31 Mar. 2028
棟朝 雅晴
Japan Society for the Promotion of Science, Grant-in-Aid for Scientific Research (C), Hokkaido University, 24K15098

集約的データ解析法による不正アクセス履歴の分析とサイバー攻撃予測への応用
科学研究費助成事業
01 Apr. 2023 - 31 Mar. 2028
南 弘征, 棟朝 雅晴, 馬場 健一, 水田 正弘, 山岡 克式
日本学術振興会, 基盤研究(B), 北海道大学, 23K28082

集約的データ解析法による不正アクセス履歴の分析とサイバー攻撃予測への応用
科学研究費助成事業
01 Apr. 2023 - 31 Mar. 2028
南 弘征, 棟朝 雅晴, 馬場 健一, 水田 正弘, 山岡 克式
日本学術振興会, 基盤研究(B), 北海道大学, 23H03392

Heuristic algorithm of exploration and exploitation to the space geometry in ciliates and amoebae
Grants-in-Aid for Scientific Research
10 Sep. 2021 - 31 Mar. 2026
中垣 俊之, 棟朝 雅晴, 田中 良巳, 國田 樹, 佐藤 勝彦
Japan Society for the Promotion of Science, Grant-in-Aid for Transformative Research Areas (A), Hokkaido University, 21H05310

Advanced mechanics of cell behavior shapes formal algorithm of protozoan smartness awoken in giorama conditions.: Algorithms Evaluation Group　　　　　　　　　　　　　　　
Grants-in-Aid for Scientific Research Grant-in-Aid for Transformative Research Areas (A)
Sep. 2021 - Mar. 2026
中垣 俊之, 佐藤 勝彦, 田中 良巳, 棟朝 雅晴, 國田 樹
Japan Society for the Promotion of Science, Grant-in-Aid for Transformative Research Areas (A), Coinvestigator, 21A402

Development and Implementation of Real World Scale Artificial Evolutionary Algorithms
Grants-in-Aid for Scientific Research
01 Apr. 2020 - 31 Mar. 2023
棟朝 雅晴
令和３年度においては、大規模かつ困難な最適化問題の解決に必要となるアルゴリズムの開発を進め、スケーラブルなリンケージ同定手法の開発ならびに実問題への適用について研究を行った。具体的には、スケーラブルなリンケージ同定手法として、sLIEM (scalable Linkage Identification with Epistasis Measures)の開発ならびに、合成人口モデルに関わる最適化問題への適用を行った。進化計算において互いに関連のある遺伝子を同定するリンケージ同定手法は、個体長の2乗オーダーの計算量を必要とし、多項式オーダーではあるが大規模問題となった場合にその計算量が課題となる。本研究課題で開発したsLIEMは、重要な変数（遺伝子）に関する摂動を中心としてリンケージ同定の対象を絞り込むことで、その計算量を削減しつつ、実問題におけるリンケージ同定の精度を確保している。
スケーラブルなリンケージ同定手法の提供例として、村田らによる合成人口モデルに関わる最適化問題へ適用し、従来手法と比較した優位性を検証した。合成人口モデルは、公開されている自治体の人口分布に関する統計データから、住民それぞれの世帯構成を推定し合成することで、プライバシーを保護しつつ、社会シミュレーションに必要とされる基礎的な人口モデルを求める手法であり、従来手法においては擬似焼き鈍し法を用いた最適化手法が提案されていた。本研究で開発したスケーラブルなリンケージ同定を導入した並列進化計算を用いることで、生成される合成人口モデルの精度を向上（誤差を低減）することができた。また、スケーラブルなリンケージ同定に加え、さらにACO (Ant Colony Optimization)の大規模並列実装、実問題への応用についても検討に着手し、次年度にその成果を公表する計画である。
Japan Society for the Promotion of Science, Grant-in-Aid for Scientific Research (C), Hokkaido University, 20K11967

最適資源選択技術に関する研究（インタークラウドを活用したアプリケーション中心型オーバーレイクラウド技術に関する研究：主たる共同研究者）　　　　　　　　　　　　　　　
CREST
Oct. 2015 - Mar. 2021
MUNETOMO Masaharu
JST, Principal investigator, Competitive research funding

Development of a communication scheme for many site coupled calculations
Grants-in-Aid for Scientific Research Grant-in-Aid for Scientific Research (C)
01 Apr. 2014 - 31 Mar. 2017
KOBAYASHI Taizo, MUNETOMO Masaharu, JITSUMOTO Hideyuki, TAKAMI Toshiya, NANRI Takeshi, MORIE Yoshiyuki
A big bottleneck in coupled simulations and large scale high performance computing is files I/O. In order to reduct this bottleneck, we have studied and developed a communication scheme of which processes are able to connect each other directly. (1) We have implemented a simple and compact communication library: NSTDIO. The interface of this communication library is designed based on "stdio" of C programing language. This library is enabled to download from GitHub https://github.com/y-morie/nstdio (2) We have developed and implemented a framework of jobs/processes cooperation on inter-site: EGCPOPS. This framework realizes jobs cooperation on inter-site without any administrator authority. This framework is also available from GitHub https://github.com/RyzeVia/exgcoup (3) Because of these framework work well, we have also studied and developed a conceptual design of mechanism for uncertainty: Self-Referential eXecutable: SRX.
Japan Society for the Promotion of Science, Grant-in-Aid for Scientific Research (C), Teikyo University, 26330146

Study on High-Performance and Reliable Load Balancing Technology for Academic Federated Cloud Platforms
Grants-in-Aid for Scientific Research Grant-in-Aid for Scientific Research (A)
31 May 2012 - 31 Mar. 2016
AIDA Kento, SAKANE Eisaku, MUNETOMO Masaharu, KOBAYASHI Taizo, Abdul-Rahman Omar
Federated cloud organized by cloud platforms distributed over multiple sites is a promising platform for research and education in academic communities. However, a load balancing technology to assign application programs to suitable cloud platforms has not been established. In this project, we conducted the study on the load balancing technology that enabled high-performance and reliable computing by utilizing cloud platforms distributed over multiple sites. We proposed application performance models on cloud platforms and developed a software system to select suitable cloud platforms by collecting application performance characteristics.
Japan Society for the Promotion of Science, Grant-in-Aid for Scientific Research (A), National Institute of Informatics, 24240006

Extremely large-scale optimization problem-solving using robust and scalable evolutionary computation
Grants-in-Aid for Scientific Research Grant-in-Aid for Scientific Research (C)
01 Apr. 2010 - 31 Mar. 2014
MUNETOMO Masaharu
We have developed a series of evolutionary algorithms such as BOA-MD, an extension of BOA introducing mixture distributions, BHCS by combining BOA with local search, ARGA and BRGA for MINLP (Mixed-Interger Non-Linear Programming). We also developed large-scale parallel evolutionary algorithms in a many core architecture and cloud computing environment.
As applications to real-world problems, we applied evolutionary algorithms to optimal resource allocation problem in cloud computing environment as a MINLP problem, de novo ligand docking problems to find promising structures of medicines automatically, and so on, to show the effectiveness of our approach.
Japan Society for the Promotion of Science, Grant-in-Aid for Scientific Research (C), Hokkaido University, Principal investigator, Competitive research funding, 22500196

アカデミックインタークラウドの実現に向けた連携基盤技術に関する研究　　　　　　　　　　　　　　　
国立情報学研究所一般公募型共同研究
2014 - 2014
棟朝 雅晴
国立情報学研究所, Principal investigator, Competitive research funding

Design and Development of Advanced IT Research Platform for Information Explosion Era
Grants-in-Aid for Scientific Research Grant-in-Aid for Scientific Research on Priority Areas
2006 - 2010
ADACHI Jun, TANAKA Katsumi, NISHIDA Toyoaki, KUNIYOSHI Yasuo, SUDOH Osamu, KUROHASHI Sadao, HARA Takahiro, MATSUOKA Satoshi, TAURA Kenjiro, TATEBE Osami, MUNETOMO Masaharu, HIROTSU Toshio, MATSUBARA Jin, SHIMOJYO Shinji, CHIBA Shigeru, YUASA Taichi, MATSUYAMA Takashi, CHIKAYAMA Takashi, KONDO Toru, KONO Kenji, OKAMOTO Masahiro, AIDA Kento, KAMADA Tomio, KITSUREGAWA Mararu, YAMANA Hayato, NAKAMURA Yutaka, KOBAYASHI Hiroaki, NAKAJIMA Hiroshi
This project implemented a common research infrastructure for all the research groups participating in this priority-area research initiative, accordingly supported all research activities in this initiative. Providing this infrastructure, we succeeded in accelerating shared utilization of research facilities and resources within the limitation of research funding and strengthening the collaboration among research groups. These shared facilities include (a)TSUBAKI: a open search engine for large-scale corpus, (b)InTrigger : Widely-distributed computing test-bed, (c)IMADE : an environment for real-world interaction measurement and analysis, and (d) prototyping for sensor-network based preventive medicine.
Japan Society for the Promotion of Science, Grant-in-Aid for Scientific Research on Priority Areas, National Institute of Informatics, Coinvestigator not use grants, Competitive research funding, 18049073

Design Optimization using Large-Scaled Parallel Evolutionary Algorithms with Gene Analysis
Grants-in-Aid for Scientific Research Grant-in-Aid for Young Scientists (B)
2006 - 2008
MUNETOMO Masaharu
本研究においては、遺伝子解析による高度な進化計算アルゴリズムを大規模並列化するとともに、グリッドコンピューティングシステムなど最新の大規模並列計算環境における実装を行った。さらには、最適化問題を解きたい設計者の有するシミュレーションプログラムとの連携を容易に行うためのフレームワークを提案することで、大規模かつ複雑な現実の設計問題、最適化問題を解決するための基盤となるシステムのプロトタイプを開発した。
Japan Society for the Promotion of Science, Grant-in-Aid for Young Scientists (B), Hokkaido University, Principal investigator, Competitive research funding, 18700219

メタ計算の進化的探索に基づく正当で効率的なプログラムの自動生成
科学研究費助成事業 萌芽研究
2004 - 2005
赤間 清, 棟朝 雅晴
メタ計算は、仕様から正当なプログラムを生成するための新しい方法である。
メタ計算では、計算状態の束をメタ節集合で表現し、メタルールで変換を次々に行うことによって新しいメタ節を得る。最初と最後のメタ節集合のペアから正当なルールを得ることができ、そのようなルールを集めることによって、プログラムを得ることができる。
得られたプログラムの(部分)正当性は、個々のメタルールの正当性から保証できる。
正しいメタルールを仕様から作る方法もすでに与えられている。
このルール生成のパラメータは、最初のメタ節集合とルール適用の2つである。ルール適用は、メタ節集合のどの位置にどのメタルールを適用するか、それを何回繰り返すかである。本研究では、これらのパラメータを遺伝子にコーディングして、進化的探索によって、よりよいルールを低コストで発見する方法を提案した。その際、ルールの評価として、分岐数が少なく、節のサイズが小さいものを優先する指標を用いた。
ポイントとなるのは遺伝子へのコーディングである。ルール適用の可能性はそのときのメタ節集合によって大きく変わるので、あらかじめ遺伝子の空間を固定するのは得策ではない。そこで、メタ節集合とメタルールが与えられたとき、ルール適用の可能性の集合を高速に計算する手法や、その中のルール適用の1つの可能性をメタ節集合に適用して、次のメタ節集合を求める手法を考案し、それらをET言語に組み込み述語として導入した。これにより、メタ計算を基礎とした進化的探索のアルゴリズムの実現が容易になり、メタ計算の探索に基づくプログラム生成を効率的に行うことが可能になった。
日本学術振興会, 萌芽研究, 北海道大学, Coinvestigator not use grants, Competitive research funding, 16650003

遺伝子解析に基づく遺伝的アルゴリズムの開発とシステム設計への応用
科学研究費助成事業 若手研究(B)
2003 - 2005
棟朝 雅晴
本研究においては、これまでに提案されたリンケージ同定に基づく遺伝的アルゴリズムをさらに発展させ、より一般的な枠組みとして、遺伝子解析に基づく遺伝的アルゴリズムを開発することを目的とし、さらにそのシステム設計問題への応用をはかっている。バイオインフォマティクスの分野では、遺伝子解析に関する手法が数多く提案されているが、それらを参考にリンケージの同定、ビルディングブロックの検出、交叉手法の改良などを行い、より高性能で高い信頼性を有する遺伝的アルゴリズムを開発する。
本年度においては、分布推定手法の改良を図るとともに、効率的な並列化手法を開発することで、大規模な問題への対応を図った。具体的には、割り当て関数による重み付けを行うことで精度を向上させ分布推定に要する計算コストを削減するための手法を開発するとともに、ベイジアンネットワークに基づく確率モデル構築を行う分布推定アルゴリズムにおいて、ネットワーク構築の効果的な並列化手法を開発した。
さらには、リンケージ同定と分布推定アルゴリズムの双方の利点を組み合わせたアルゴリズムD^5の開発を行い、問題規模の増大に対するスケーラビリティに優れた手法を開発した。さらに、D^5の実数値問題への適用についても議論するとともに、その性能について評価を行った。
提案手法の適用例として、特に、蛋白質の構造エネルギー最小化問題へ、並列化した分布推定による探索手法を適用することで、本研究で開発した手法の有効性を検証している。
日本学術振興会, 若手研究(B), 北海道大学, Principal investigator, Competitive research funding, 15700175

遺伝的アルゴリズムの設計理論の確立およびその大規模分散システム設計への応用
科学研究費助成事業 若手研究(B)
2001 - 2002
棟朝 雅晴
本研究においては、線型・非線型および単調・非単調性を調べることによって関連する遺伝子座のまとまりであるリンケージ集合を同定し、効果的な交叉を実現する手法に関してその理論的基礎を確立するとともに、理論に基づいた遺伝的アルゴリズムの設計、特に交叉手法の改良を行うことで、大規模な最適化問題におけるアルゴリズムの性能を評価した。リンケージ集合の構造として、これまでの研究では比較的単純な線形の構造が仮定されていたが、本研究では階層構造を有するリンケージの同定を行う手法である、hLIEM(hierarchical Linkage Identification with Epistasis Measure)およびhLIEM^2(hierarchical Linkage Identification with Epistasis Measure considering Monotonicity)を開発した。これら手法においては、小さなリンケージ集合をまとめてより大きなリンケージ集合を生成するという過程を繰り返すことで、現実問題の多くが有すると考えられる問題の階層構造を同定することが可能となり、交叉など遺伝的操作をさらに効率的に適用することが可能となった。
本研究で開発した階層型のリンケージ同定アルゴリズムの有効性を、階層型のトラップ関数に代表される各種のテスト関数において確認したが、さらに現実の問題として都市圏ネットワーク設計問題へと適用し、その有効性を確かめた。大規模なネットワークを設計する場合、ゆるやかな階層構造を有するようなネットワークとして設計を行うことが考えられるが、本研究で提案するアルゴリズムにより、人間の設計者が介在することなく、計算機により自動的に適切な階層構造を有するネットワークを構成することができ、従来の手法に比べて、より低コストで高性能なネットワークを設計することができた。
日本学術振興会, 若手研究(B), 北海道大学, Principal investigator, Competitive research funding, 13780182

Construction of Program Synthesis System Based on Equivalent Transformation
Grants-in-Aid for Scientific Research Grant-in-Aid for Scientific Research (B)
2000 - 2002
AKAWA Kiyoshi, MUNETOMO Masaharu, MIZUTA Masahiro
1. Development of a general theory of meta-computation We developed a general theory for program synthesis based on a new class of representation called separated descriptions. We clarified the fundamental difference between program transformation and fule generation, and showed that program transformation can be regarded as a part of our program synthesis.
2. Extension of the program sysnthesis system using first-order expressions We constructed an expertimental program synthesis system, which automatically synthesizes a program by repeatedly generating and accumulating new rules. We obtained several successful results. For example, the system can automatically synthesize, from a definition of a language represented by a first-order expression, a program corresponding to a finite automaton that recognizes the language.
3. Extension of meta-computation We extended meta-decriptions to improve the system to be able to generate a larger class of rules. We introduced expressions of constraints on variables and extended a theory of meta-computation. This enabled the system to generate with applicability conditions and can be more controllable. We also investigated a theoretical foudation of generation of such rules.
4. Evaluation of the program synthesis system We applied the system to generation of parser programs. Highly optimized programs are obtained at the experiments. However they can be more efficient by optimization of execution parts of rules and by transforming the rules into deterministic ones. Although we conducted them manually, complete automation is a future work.
Japan Society for the Promotion of Science, Grant-in-Aid for Scientific Research (B), HOKKAIDO UNIVERSITY, Coinvestigator not use grants, Competitive research funding, 12480076

通信資源管理への人工進化モデルの適用とその大規模ネットワークにおける評価
科学研究費助成事業 奨励研究(A)
1999 - 2000
棟朝 雅晴
本研究では、進化的計算手法、特に遺伝的アルゴリズムを用いた経路制御アルゴリズムに関して、大規模ネットワーク上での性能評価を行い、アルゴリズムの改良を行った。改良点としては、適応度を評価するために必要なネットワーク状態の観測法について、従来は評価パケットが到着するまでの時間を用いていたのに対し、改良したアルゴリズムでは各リンクの負荷状態を離散化した情報を収集することでより高速かつ効率的にネットワークの負荷状態を観測することができた。さらには代替経路を生成するために使用される遺伝的操作に関してもリンクの負荷状態を考慮した交叉および突然変異の手法を開発することでアルゴリズムの改良を行った。提案したアルゴリズムの評価を行うため、イベントシミュレーション手法に基づくネットワークシミュレータを開発し、大規模シミュレーションにより改良手法の有効性を確認した。
本研究では、さらに、ネットワーク資源管理アルゴリズムとして分散型帯域幅割当アルゴリズムを開発した。特定の通信リンクや小規模ネットワークを対象とした場合、通信帯域割当などの資源管理は集中管理方式により比較的容易に行うことができるが、大規模ネットワーク上においては、必然的に不確実な情報を基にした分散管理を行わざるを得なくなり、単純な最適化問題としてとらえることはできない。このような状況において、遺伝的アルゴリズムを用いた分散型帯域幅割当アルゴリズムを構築し、その有効性をシミュレータによるシミュレーション実験と通して検証した。
日本学術振興会, 奨励研究(A), 北海道大学, Principal investigator, Competitive research funding, 11780179

進化的計算手法を用いた適応型経路制御アルゴリズムの構築
科学研究費助成事業 奨励研究(A)
1997 - 1998
棟朝 雅晴
本研究では、複雑系、人工生命に関連して近年注目を集めている遺伝的アルゴリズムをインターネットにおける経路制御アルゴリズムへ応用することで、ネットワークの負荷変化に動的するアルゴリズムを構築する。
現在一般に使用されているルーティングアルゴリズムでは、ネットワークの負荷状態や通信リンクの性能などを考慮せず、単純に通過したゲートウェイの数で表されるホップカウント距離を用いていた。そこで、ネットワークの負荷状態を考慮した適応型のルーティングアルゴリズムを少ない通信負荷で実現するため、遺伝的アルゴリズムにより必要な代替経路群を動的に生成するアルゴリズムを構築した。
代替経路を効率的に生成するため、特別に設計された経路遺伝的操作(Path Genetic Operators)を提案した。提案するルーティングアルゴリズムにおいては、初期経路として、従来と同様のホップカウント距離に基づいた最短経路を使用し、ある一定回数以上使用される経路に関して、代替経路を動的に生成し、各代替経路の通信遅延に応じてパケットを分配する。通信遅延の観測に関しては、ルーティングテーブルに存在する各経路に関して、定期的に通信遅延の観測パケットを送出するが、ルーティングテーブルには頻繁に使用される少数の代替経路のみ存在するため、少ない通信負荷で効果的な観測が行われる。
アルゴリズムの評価を行うため、離散事象シミュレータを構築し、その上でシミュレーション実験を行い、従来の代表的な手法であるRIP,SPFと比較した。その結果、今回提案した基本的なアルゴリズムにより、ネットワークの負荷が減少し、効果的に通信パケットが送出されていることが確かめられた。
日本学術振興会, 奨励研究(A), 北海道大学, Principal investigator, Competitive research funding, 09780225