Kenji FUKUMIZU
Professor, Intelligent Information Processing Group, Department of Statistical Modeling
Phone: +81-50-5533-8540
E-mail: fukumizu
Personal
Homepage:
http://www.ism.ac.jp/~fukumizu/
Date of Birth: May, 1966
Education: Bachelor of Science, Dept. of Science, Kyoto Univ., 1989.
Ph.D. of Science, Dept. of Science, Kyoto Univ., 1996.
Degree: Ph.D. of Science, Kyoto Univ., 1996, Title of Ph.D Thesis:A regularity condition of the information matrix of a multilayer perceptron network.
Employment: 1989.4-1997.12     Research and Development department in Ricoh Co., Ltd. (Japan) Researcher.
1998.1-2000.1       Lab for Information Synthesis, Brain Science Institute, RIKEN (Japan) Research scientist.
2000.2- 2005.3      Institute of Statistical Mathematics, (Japan) Associate professor.
2005.4- 2009.9                Department of Statistical Modeling, Institute of Statistical Mathematics, (Japan) Associate professor.
2009.10-                 Department of Statistical Modeling, Institute of Statistical Mathematics, (Japan) Professor.
Honors &
Awards:
Research award of Japanese Neural Network Society, 1997, "Active learning - optimal queries -". Young researcher award, of Japanese Neural Network Society, 1998, "Special statistical property of multilayer neural networks". Paper award of Japanese Neural Network Society, 2001, Fukumizu, K. and Amari, S. "Local Minima and Plateaus in Hierarchical Structures of Multilayer Perceptrons", Neural Networks, 13 (3) (2000), 317-327.
Research
Fields:
Machine Learning, Statistical learning theory
Research
Themes:
Statistical methods with positive definite kernels, Probabilistic inference on graphs, Statistical estimation theory of unidentifiable models

Selected Papers
(1) Sriperumbudur, B., Fukumizu, K., Gretton, A., Lanckriet, G., and Scholkopf, B. (2010): Kernel choice and classifiability for RKHS embeddings of probability distributions. Advances in Neural Information Processing Systems 22. to appear.
(2) Gretton, A., Harchaoui, Z., Fukumizu, K., Sriperumbudur, B. (2010): A fast, consistent Kernel two-sample test. Advances in Neural Information Processing Systems 22. to appear.
(3) Watanabe, Y., and Fukumizu, K. (2010): Graph zeta function in the Bethe free energy and loopy belief propagation. Advances in Neural Information Processing Systems 22. to appear.
(4) Watanabe, Y., and Fukumizu, K. (2009): Loop series expansion with propagation diagrams Journal of Physics A: Mathematical and Theoretical 42, 045001, 18pp.
(5) Fukumizu, K. (2008): Exponential manifold by reproducing kernel Hilbert spaces. in Algebraic and Geometric mothods in statistics (P.Gibilisco, E.Riccomagno, M.-P.Rogantin, and Henry Winn eds.) Cambridge University Press, 291-306
(6) Fukumizu, K. F.R. Bach and M. I. Jordan. (2008): Kernel dimension reduction in regression. The Annals of Statistics, 37(4), 1871-1905.
(7) Fukumizu, K., Sriperumbudur, B., Gretton, A., and Scholkopf, B. (2008): Characteristic Kernels on groups and semigroups. Advances in Neural Information Processing Systems 21, 473-480
(8) Fukumizu, K., A. Gretton, X. Sun, and B. Schoelkopf (2008): Kernel Measures of Conditional Dependence. Advances in Neural Information Processing Systems 21, 489-496.
(9) Gretton, A. Fukumizu, K. Teo,C.H., Song, L., Schoelkopf, B., and Smola, A. (2008): A Kernel Statistical Test of Independence. Advances in Neural Information Processing Systems 21, 585-592.
(10) Fukumizu, K. (2008): Statistical estimation with positive definite kernels. in Statistical Science in the 21th Century III (G. Kitagawa and A. Takemura eds.), 257-290. University of Tokyo Press. (in Japanese)
(11) Shiraishi, Y. and Fukumizu, K. (2008): Game theoretical combination of binary classifiers for multi-class classification. IEICE Transactions on Information and Systems Vol.J91-D No.6, 1528-1537 (in Japanese)
(12) Fukumizu, K., F. R. Bach and A. Gretton. (2007): Statistical consistency of kernel canonical correlation analysis, Journal of Machine Learning Research, 8, 361-383.
(13) Marco Cuturi, Kenji Fukumizu, and Jean-Philippe Vert (2005): Semigroup Kernels on Measures, Journal of Machine Learning Research, 6, 1169-1198.
(14) Watanabe, S., Fukumizu, K., Hagiwara, K., and Amari, S. (2005): Learning Theory of Singular Models, The IEICE Transactoins J88-D2 (2), 159-169 (in Japanese)
(15) Fukumizu, K., Bach, F.R., and Jordan, M.I. (2004): Dimensionality reduction for supervised learning with reproducing kernel Hilbert spaces, Journal of Machine Learning Research, 5, 73-99.
(16) Fukumizu, K. and Akaho, S. and Amari, S. (2003): Critical lines in symmetry of mixture models and its application to component splitting, Advances in NIPS, 15, 865-872.
(17) Fukumizu, K. and Kuriki, S. (2003): Statistical inference in singular models - an approach by tangent cone -, Journal of Japan Neural Network Society, 10 (4), 201-210. (in Japanese)
(18) Fukumizu, K. (2003): Likelihood ratio of unidentifiable models and multilayer neural networks, Annals of Statistics, 31 (3), 833-851.
(19) Fukumizu, K. (2001): Geometry of neural networks: Natural gradient for learning, Handbook of Biological Physics, vol.4: Neuro-informatics and Neural Modelling, Elesvier, 731-769.

(20) Fukumizu, K. and Amari, S. (2000): Local minima and plateaus in hierarchical structures of multilayer perceptrons, Neural Networks, 13 (3), 317-327.
(21) Fukumizu, K. (2000): Statistical active learning in multilayer perceptrons, IEEE Trans. Neural Networks, 11 (1), 17-26.
(22) Amari, S., Park, H., and Fukumizu, K. (2000): Adaptive method of realizing natural gradient learning for multilayer perceptrons, Neural Computation, 12 (6), 1399-1409.
(23) Fukumizu, K. (1999): Generalization error of linear neural networks in unidentifiable cases, Lecture Notes in Artificial Intelligence 1720, Algorithmic Learning Theory, Springer Verlag, 51-62.
(24) Fukumizu, K. (1997): Active learning in neural networks, Journal of IPSJ, 8 (7), 569-574 (in Japanese).
(25) Fukumizu, K. (1996): A regularity condition of the information matrix of a multilayer perceptron network, Neural Networks, 9 (5), 871-879.
(26) Ishii, S., Watanabe, S. and Fukumizu, K. (1996): A network of chaotic elements for information processing, Neural Networks, 9 (1), 25-40.
(27) Fukumizu, K. (1996): Active learning in multilayer perceptrons, Advances in NIPS, 8, 295-301.
(28) Fukumizu, K. and Watanabe, S. (1996): Optimal training data and predictive error of polynomial approximation, Trans. IEICE, Vol.J79-A(5), 1100-1108 (in Japanese).
(29) Watanabe, S. and Fukumizu, K. (1995): Probabilistic design of layered neural networks based on their unified framework, IEEE Trans. Neural Networks, 6 (3), 691-702.

Selected Publications
(1) Neural Network Systems Techniques and Applications 1: Algorithms and Archtechtures, Academic Press (1997).
(2) Statistical Theory of Singular Models, Iwanami Shoten (2004). (in Japanese)
(3) Theory and Practice of Learning Systems, Morikita-Shuppan. (2005). (in Japanese)

Membership of Academic Societies
The Institute of Electronics, Information and Communication Engineers; Japanese Neural Network Society; The Institute of Mathematical Statistics

Professional Services
Neural Networks, Associate editor. Foundations and Trends in Machine Learning, Editorial Board. Annals of the Institute of Statistical Mathematics, Associate editor. Technical group on Information-based Induction Sciences. The Institute of Electronics. Information and Communication Engineers, Commissioner.

HOME