Researcher Database

Researcher Profile and Settings

Master

Affiliation (Master)

  • Faculty of Economics and Business Division of Modern Economics and Management Economic Analysis

Affiliation (Master)

  • Faculty of Economics and Business Division of Modern Economics and Management Economic Analysis

researchmap

Profile and Settings

Profile and Settings

  • Name (Japanese)

    Suzukawa
  • Name (Kana)

    Akio
  • Name

    201501033390367210

Achievement

Research Interests

  • 打ち切りデータ   生存時間データ   打切りデータ   打ち切り   正準相関分析   判別分析   生存関数   セミパラメトリック統計モデル   競合危機   セミパラメトリックモデル   生存時間   漸近理論   2変量生存関数   生存時間分布   医学データ   コピュラ   セミパラメトリック   コックス回帰分析   打切り   多変量データ   ハザード関数   差分方程式   関数クラスター分析   曲線当てはめ   関数データ   関数解析   高速計算   関数回帰分析   差分   時系列データ   

Research Areas

  • Informatics / Statistical science

Research Experience

  • 2013/04 - Today Hokkaido University
  • 2001/10 - 2013/03 Hokkaido University
  • 1998/04 - 2001/09 Obihiro University of Agriculture and Veterinary Medicine
  • 1997/04 - 1998/03 Hokkaido University Graduate School of Engineering
  • 1995/04 - 1997/03 Tokyo University of Science

Awards

  • 2005/09 日本統計学会 小川研究奨励賞
     打切りデータに基づく不偏推定 
    受賞者: 鈴川晶夫

Published Papers

  • Koshiro Yonenaga, Akio Suzukawa
    THEORY OF PROBABILITY AND MATHEMATICAL STATISTICS 108 209 - 224 0094-9000 2023/05 [Refereed]
     
    We consider the distribution of the product of a Wishart matrix and a normal vector with uncommon covariance matrices. We derive the stochastic repre-sentation which reduces the computational burden for the generation of realizations of the product. Using this representation, the density function and higher order mo-ments of the product are derived. In a numerical illustration, we investigate some properties of the distribution of the product. We further suggest the Edgeworth type expansions for the product, and we observe that the suggested approximations provide a good performance for moderately large degrees of freedom of a Wishart matrix.
  • Bayesian estimation for misclassification rate in linear discriminant analysis
    Koshiro Yonenaga, Akio Suzukawa
    JAPANESE JOURNAL OF STATISTICS AND DATA SCIENCE 4 (2) 861 - 885 2021/04 [Refereed]
  • Estimators of Bivariate Extreme Value Copulas
    Akio Suzukawa
    Annals of Public Policy, 20 (1) 117 - 144 2019/10 [Not refereed][Not invited]
  • Product of inverse Wishart and normal distributions
    Koshiro Yonenaga, Akio Suzukawa
    Proceedings of Korea and Hokkaido University 5th Workshop in Statistics 1 (1) 5 - 10 2018/02 [Not refereed]
  • The mediator effect of personality traits on the relationship between childhood abuse and depressive symptoms in schizophrenia
    Ryo Okubo, Takeshi Inoue, Naoki Hashimoto, Akio Suzukawa, Hajime Tanabe, Matsuhiko Oka, Hisashi Narita, Koki Ito, Yuki Kako, Ichiro Kusumi
    Psychiatry Research 257 (1) 126 - 131 2017/01 [Refereed]
  • Semiparametric models for lifetime distribution
    Akio Suzukawa
    Proceedings of Hokkaido University and Korea University 4th Workshop in Statistics 2016/06
  • Bivariate discriminant analysis based on Gumbel copula
    Yuki Yasuda, Akio Suzukawa
    Proceedings of Hokkaido University and Korea University 4th Workshop in Statistics 1 (1) 15 - 20 2016/05 [Not refereed]
  • Generalized Archimedean copulas
    Akio Suzukawa
    Proceedings of the 9th Conference of the Asian Regional Section of the IASC 2015/12
  • Support vector machine based on data depth
    Zhaojun Jian, Akio Suzukawa
    Economic Journal of Hokkaido University, Discussion Paper Series A (282) 1 - 15 2015/03 [Not refereed]
  • On the nonlinear projection trick in the kernel methods
    Shogo Omote, Akio Suzukawa
    Proceedings of Korea University and Hokkaido University 3rd Workshop in Statistics 2015/02 [Not refereed]
  • Reversible Jump Markov Chain Monte Carlo Method
    Ryo Yawata, Akio Suzukawa
    Proceedings of Korea University and Hokkaido University 3rd Workshop in Statistics 2015/02
  • A generalization of the shared frailty models for multivariate lifetime distributions
    Akio Suzukawa
    Proceedings of Korea University and Hokkaido University 3rd Workshop in Statistics 2015/02 [Not refereed]
  • A survival model with epidemic effects
    Akio Suzukawa
    Proceedings of Hokkaido University-Korea University Second Joint Workshop in Statistics 2013/06 [Not refereed]
  • A Class of Nonparametric Estimators for Bivariate Extreme Value Copulas
    Akio Suzukawa
    Graduate School of Economics and Business Administration, Hokkaido University Discussion paper series A (230) 2010 [Not refereed]
  • An Approximate Likelihood Procedure for Competing Risks Data
    Akio Suzukawa
    Journal of the Japan Statistical Society, 40 239 - 263 2010 [Refereed]
  • Effects of social relationships on mortality of the elderly: How do the influences changes with the passage of time?
    T. Sato, R. Kishi, A. Suzukawa, N.Horikawa, Y. Saijo, E. Yoshioka
    Archives of Gerontology and Geriatrics 47 (3) 327 - 349 2008 [Refereed]
  • Unbiased Estimation of Functionals Under Random Censorship
    Akio Suzukawa
    Journal of the Japan Statistical Society, 34 (4) 153 - 172 2004 [Refereed]
  • Semiparametric estimation based on parametric modeling of the cause-specific hazard ratios in competing risks
    Akio Suzukawa, Nobuhiro Taneichi
    Journal of Multivariate Analysis 887 (1) 80 - 100 2003 [Refereed]
  • Asymptotic properties of Aalen-Johansen integrals for competing risks data
    Akio Suzukawa
    Journal of the Japan Statistical Society 32 (1) 77 - 93 2002 [Refereed]
  • Redundancy index in canonical correlation analysis with linear constraints
    Akio Suzukawa, Nobuhiro Taneichi
    Measurement and Multivariate Analysis (S.Nishisato, Y.Baba, H.Bozdogan and K.Kanefuji Eds.), Springer 125 - 132 2002 [Refereed]
  • A formula for normalizing transformation of some statistics
    N.Taneichi, Y.Sekiya, A.Suzukawa, H.Imai
    Communications in Statistics, Theory and Methods 31 (2) 163 - 179 2002 [Refereed]
  • A Suzukawa, H Imai, Y Sato
    ANNALS OF THE INSTITUTE OF STATISTICAL MATHEMATICS 53 (2) 262 - 276 0020-3157 2001/06 [Refereed][Not invited]
     
    This paper is intended as an investigation of parametric estimation for the randomly right censored data. In parametric estimation, the Kullback-Leibler information is used as a measure of the divergence of a true distribution generating a data relative to a distribution in an assumed parametric model M. When the data is uncensored, maximum likelihood estimator (MLE) is a consistent estimator of minimizing the Kullback-Leibler information, even if the assumed model M does not contain the true distribution. We call this property minimum Kullback-Leibler information consistency (MKLI-consistency). However, the MLE obtained by maximizing the likelihood function based on the censored data is not MKLI-consistent. As an alternative to the MLE, Oakes (1986, Biometrics, 42, 177-182) proposed an estimator termed approximate maximum likelihood estimator (AMLE) due to its computational advantage and potential for robustness. We show MKLI-consistency and asymptotic normality of the AMLE under the misspecification of the parametric model. In a simulation study, we investigate mean square errors of these two estimators and an estimator which is obtained by treating a jackknife corrected Kaplan-Meier integral as the log-likelihood. On the basis of the simulation results and the asymptotic results, we discuss comparison among these estimators. We also derive information criteria for the MLE and the AMLE under censorship, and which call be used not only for selecting models but also for selecting estimation procedures.
  • Kaplan-Meier積分に対する自己一致方程式とその応用
    鈴川 晶夫, 種市 信裕
    計算機統計学 13 (2) 93 - 103 2001/05 [Refereed]
  • Taneichi Nobuhiro, Sekiya Yuri, Suzukawa Akio
    Journal of the Japan Statistical Society THE JAPAN STATISTICAL SOCIETY 31 (2) 207 - 224 1882-2754 2001/01 [Refereed]
     
    On the goodness-of-fit test for multinomial distribution, Zografos et al.(1990)proposed the φ-divergence family of statistics, which includes the power divergence family of statistics as a special case. They showed that under null hypothesis, the members of the φ-divergence family of statistics all have an asymptotically equivalent chi-square distribution. Furthermore, Menendez et al.(1997)derived an asymptotic expansion for the null distribution of the φ-divergence statistic. In this paper, we derive an approximation for the distribution of the φ-divergence statistic under local alternatives. The approximation is based on the continuous term of the asymptotic expansion for the distribution of the φ-divergence statistic. By using the approximation, we propose a new approximation for the power of the statistic. The results are generalizations of those derived by Taneichi et al.which discussed the power divergence statistic. We numerically investigate the accuracy of the approximation when two types of concrete φ-divergence statistics are applied. By the numerical investigation, we find that the present approximation performs better than the other approximations.
  • ESTIMATION OF SURVIVAL FUNCTION BASED ON MODELING OF CENSORING PATTERN
    Suzukawa Akio, Taneichi Nobuhiro
    Journal of the Japan Statistical Society 30 (2) 177 - 195 2000/12 [Refereed]
  • STATISTICAL INFERENCE IN A CANONICAL CORRELATION ANALYSIS WITH LINEAR CONSTRAINTS
    Akio Suzukawa
    STATISTICAL INFERENCE IN A CANONICAL CORRELATION ANALYSIS WITH LINEAR CONSTRAINTS 27 (1) 93 - 107 1997/06 [Refereed]
  • LINEAR HYPOTHESIS TESTING IN A RANDOM EFFECTS GROWTH CURVE MODEL
    Suzukawa Akio, Sato Yoshiharu
    Journal of the Japan Statistical Society 26 (2) 135 - 143 1996/12 [Refereed]
  • THE NON-NULL DISTRIBUTION OF THE LIKELIHOOD RATIO CRITERION FOR ADDITIONAL INFORMATION HYPOTHESIS IN CANONICAL CORRELATION ANALYSIS
    Suzukawa Akio, Sato Yoshiharu
    JOURNAL OF THE JAPAN STATISTICAL SOCIETY 26 (1) 91 - 100 1996 [Refereed]
  • SUZUKAWA Akio, SATO Yoshiharu
    Journal of the Japan Statistical Society THE JAPAN STATISTICAL SOCIETY 25 (1) 71 - 80 0389-5602 1995/06/01 
    Assuming that X=(X(1)', X(2)', …, X(q)')' is distributed according to Np(μ, Σ) and each X(j) has pj components, where p=p1+…+pq and p1p2≥…≥pq, we consider the testing of the hypothesis that for a=2, …, q, all the smallest pa-ka canonical correlations between (X(1)', X(2)', …, X(a-1)')' and X(a) are zero. In the case of k2=k3=…=kq=0, this hypothesis is equivalent to the hypothesis of independence of X(1), X(2), …, X(q). This testing problem can then be considered to be a generalization of the problem of testing independence. In this paper we drive the likelihood ratio statistic (LR statistic) and obtain the asymptotic expansion of its null distribution. We also obtain the asymptotic expansions of the null distributions of the other two statistics, which are the statistics corresponding to Lawley-Hotelling's trace and the Bartlett-Nanda-Pillai's trace criterion.
  • Comparison of powers and exact Bahadur slopes of some tests for dimensionality in MANOVA models
    Akio Suzukawa, Yoshiharu Sato
    Communications in Statistics : Theory and Method 23 (12) 3581 - 3591 1994 [Refereed]

MISC

Books etc

  • 確率と統計 ―基礎と応用―
    木村 俊一, 古澄 英男, 鈴川 晶夫 (Joint work第7章, 第8章, 第9章)
    朝倉書店 2003/09 (ISBN: 9784254111026)

Research Projects

  • Japan Society for the Promotion of Science:Grants-in-Aid for Scientific Research
    Date (from‐to) : 2016/04 -2019/03 
    Author : Kishi Reiko, Kajiwara Junboku, Sata Fumihiro, Saijo Yasuaki, Yoshioka Eiji
     
    To investigate the effects of pre-and postnatal exposure to environmental chemicals on child’s neurodevelopment, puberty onset, and allergies, we evaluated these adverse health impacts in terms of single and multiple exposures, as well as gene-environment interaction. A significant decreasing secular trend of urinary MEHP concentration of 7 years old children was found in the sample collected between 2012 and 2017. Prenatal exposures to MECPP was associated with increase a risk of behavioral problems among 5 years old children and MnBP, MiBP, MEHP were associated with early puberty onset among girls. Prenatal exposure to BPA did not have any associations with neurodevelopment or puberty onset. Prenatal exposure to multiple chemicals disrupted sex hormone levels in cord blood, especially MEHP and PFOS were highly contributions to hormone levels. A significant decreasing 2D:4D ratio was found in cord blood XbaI polymorphism of estrogen receptor 1 gene.
  • Japan Society for the Promotion of Science:Grants-in-Aid for Scientific Research
    Date (from‐to) : 2013/04 -2018/03 
    Author : Suzukawa Akio
     
    Multivariate survival time distribution was modeled using copula (a function that connects multidimensional probability distribution and its marginal distributions). Based on this model, we developed a discriminant analysis method of multivariate survival time data.In particular, focusing on the application range to medical data, we developed discriminant analysis methods applicable to actual multivariate survival time data in medical research In data analysis of multivariate survival time, censoring of data and non-multivariate normality are important problems. We proposed a parametric model for censoring patterns. Based on copula modeling of multivariate survival distributions, we developed flexible discriminant analysis methods.
  • 日本学術振興会:科学研究費助成事業
    Date (from‐to) : 2005 -2007 
    Author : 鈴川 晶夫
     
    多変量競合危険データ解析のためのセミパラメトリックモデルを構築し,そのモデルに基づく分析手法を開発した。また,癌の再発時間データや透析患者のカテーテル感染時間データなど,様々な実際の医学データにそれを適用することによって,その医学データへの応用上の有効性や問題点を明らかにした。理論面では,開発ざれたセミパラメトリック手法が既存のノンパラメトリック手法に比ベてどの程度効率的であるかを明らかにした。漸近相対効率を計算することにより,大標本の場合において,セミパラメトリック推定法が非常に効率的であることが示された。また、シミューレーションによる推定誤差評価によって小標本の場合においても,セミパラメトリック推定法の推定誤差がノンパラメトリック推定法に比べて小さいことが示された。 また、医学データへの応用面では,セミパラメトリックモデルの妥当性や原因別ハザード関数や変量間の関連性(例えば,癌の1度目の再発時間と2度目の再発時間の再発部位別関連性など)をどのように評価・表示すべきであるかについて統計学と医学の両方の立場から検証した。具体的なモデルとしては,最初に起こる関心イベントまでの時間(例えば,大腸癌の術後競合危険の場合には,大陽癌再発時間と肝臓への転移時間とその他部位への転移時間のミニマム)を固定した条件のもとで,それぞれのイベントの条件付確率に対してのみ,多項ロジスティックモデルなどのパラメトリックモデルを導入することを提案した。臨床試験などで得られる競合危険データに対して,このモデルに基づく分析法を適用した。その結果,競合するいくつかのリスクのもとでの生存時間と,それに影響を与える共変量(例えば,生活習慣など)の関係について,医学的見地からも有用な知見が得られることが示された。
  • Japan Society for the Promotion of Science:Grants-in-Aid for Scientific Research
    Date (from‐to) : 2002 -2005 
    Author : MIZUTA Masahiro, SATO Yoshiharu, MURAI Tetuya, SUZUKAWA Akio, MINAMI Hiroyuki, KOMIYA Yuriko
     
    In most conventional data analysis methods, we assume that data set is regarded as a set of numbers with some structures, for example a set of vectors or a set of matrices etc. Nowadays, we must often analyze more complex data. One type of the complex data is functional data structure ; data themselves are represented as functions. Ramsay and Silverman have studied function data analysis (FDA) as the analysis method to function data since the 1990's. They have published excellent books on FDA (Ramsay & Silverman, 1997, 2002). In this study, we promoted the research of the function data analysis method, the building of a theory system, an application study to the practical problems and so on. Specifically, we discussed with Prof. Ramsay and the experts of the statistical science and the information engineering. Based on those discussions, we reviewed about the expansion of the coverage of FDA. Moreover, by using the knowledge of the nonlinear data analysis, the new analysis methods were developed. Study results are classified into the following. (1)The definition of the framework of the function data : We investigated a research trend about the function data analysis. We reviewed the directionality of this research task. (2)Developments of methods for Functional Data Analysis : We developed methods for FDA including functional regression, functional MDS and functional clustering. (3)Discrete functional data analysis : We proposed a method to find the structures that the discrete functional data have by utilizing the proposed high order differences. (4)The research of the related field : We dealt with virtual parallel computer environment, dimension reduction methods and variable selection as the related topics.
  • 日本学術振興会:科学研究費助成事業
    Date (from‐to) : 2002 -2004 
    Author : 鈴川 晶夫
     
    本研究は、打ち切りのある2変量生存時間データ解析のためのセミパラメトリックモデルを構築し、そのモデルに基づく分析手法を開発することを目的とした。また、癌の再発時間データや透析患者のカテーテル感染時間データなど、様々な実際の医学データにそれを適用することによって、その医学データへの応用上の有効性や問題点を明らかにすることを目的とした。 本年度は、前年度までに行った理論的(漸近論的)推測効率評価を精密化し、推測効率に関する数値的検討を行った。応用上は、必ずしも十分大きな標本数が得られるとは限らないため、開発手法が実用的なものとなるためには、小標本のもとでの推測効率を数値的に検討することは非常に重要である。また、構築されたセミパラメトリックモデルの妥当性に関して、医学データ解析の立場から検討した。実際の医学データに開発手法を適用し、その応用上の問題点などを明らかにするとともに、手法の改良・拡張を行った。具体的には、以下の項目を遂行した。 〓理論的効率評価の精密化 セミパラメトリック推定量のノンパラメトリック推定量に対する漸近相対効率を精密に計算することにより、小標本のもとで、どのようなモデルを導入することにより、どの程度効率が改善されるかを明らかにした。 〓数値シミュレーションによる推測法の効率比較 2変量指数分布などを仮定した数値シミュレーションにより、推定量の推定誤差などを数値的に評価し、様々な推定法の望ましさに関して比較検討した。医学データへの応用 実際の医学データに対して開発手法を適用し、分析結果の解釈の仕方やその表示法について検討を行った。特に、2変量生存関数や変量間の従属性の視覚的に表示法を開発した。なお、ここで扱う医学データは、医学・統計学関連の学術雑誌・図書において既に公表されているものであり、被験者の同意や社会的コンセンサスを必要とするものではない。 〓医学的立場からの有効性の検証 分析結果の解釈・表示法を含めて、開発手法の有効性や問題点を医学的見地から検証した。
  • Japan Society for the Promotion of Science:Grants-in-Aid for Scientific Research
    Date (from‐to) : 1997 -2000 
    Author : MIZUTA Masahiro, SUZUKAWA Akio, MURAI Tetuya, SATO Yoshiharu, IMAI Hideyuki, MINAMI Hiroyuki
     
    The aim of the research is to construct a series of methods for nonlinear data analysis. Data analysis methods with algebraic curve fitting are fundamental methods in the research. We had studied algebraic curve fitting at first. We developed an algorithm to derive the exact distance between a point and algebraic curve. A method for finding an algebraic curve structures in data is proposed. According to the progress of the research, we also study Generalized Principal Components Analysis and Principal Curves. For related data analysis methods, Sliced Inverse Regression, Functional Data Analysis and Principal Points are studied. Sliced Inverse Regression with Projection Pursuit is proposed and numerical examples show the advantages of the proposed method. A method of nonlinear functional regression analysis is developed.
  • 日本学術振興会:科学研究費助成事業
    Date (from‐to) : 1998 -1999 
    Author : 鈴川 晶夫
     
    本研究において、打切りのある多変量データに対する正準相関分析法を開発し、様々な実データにそれを適用することによって、その有効性を明らかにした。正準相関分析法を回帰分析法の一般化(一方の変量集合が一変量のとき、線形回帰分析に帰着するという意味において)と捉えることによって、打切りのある多変量データに対する正準相関分析法の開発を行なった点が本研究の大きな特色である。生存時間データに対する回帰分析法としては、線形回帰分析と共にコックス回帰分析がよく用いられる。そこで、本研究において、線形回帰分析の一般化としての正準相関分析とコックス回帰分析の一般化としての正準相関分析法の開発を行った。 線形回帰分析の一般化としての正準相関分析を開発することの本質は、打切りデータに基づいて、2変量間の共分散構造の推定法を開発することにある。そこで、カプラン・マイヤー積分統計量に基づく推定量とその性質を明らかにし、情報量基準AICの打切りデータへの拡張を行った(鈴川、種市(2000)、Suzukawa, Imai and Sato (2000))。 コックス回帰分析の重要性は、生存時間と共変量の間の従属性をハザード(hazard)という概念を通じて解釈可能であることにある。本研究において、コックス回帰分析の一般化としての正準相関分析法の開発を行った。この分析法により、複数の共変量が複数の生存時間のハザードに与える影響を総合的に評価・解釈可能であることが、実データへの適用結果やシュミレーション結果により確認された(Suzukawa and Taneichi(1999))。


Copyright © MEDIA FUSION Co.,Ltd. All rights reserved.