Japanese version is also available. 
picture

Kazumi Saito

[Contact] 2-4, Hikaridai, Seika-cho,
Soraku-gun, Kyoto 619-0237 Japan
TEL: +81 774 93 5137, FAX: 0774-93-5155
E-MAIL: saito@cslab.kecl.ntt.co.jp

Profile

Kazumi Saito received the B.S. degree from Keio University in 1985 and the PhD degree from University of Tokyo in 1998. In 1985, he joined the Electrical Communication Laboratories, NTT. Currently, he is a senior research scientist in NTT Communication Science Laboratories.

Papers

Refereed Journal Papers

  1. Naonori Ueda and Kazumi Saito, "Parametric Mixture Models for Multi-Topic Text (in Japanese), "IEICE Trans., Vol.J87-D-II, No.3, pp.872--883, 2004.
  2. Naonori Ueda and Kazumi Saito, "Multi-topic Text Model for Topic-based Text Retrieval (in Japanse)," Trans. of IPSJ, Vol.44, No.14, pp.1--8, 2003.
  3. Masahiro Kimura, Kazumi Saito, and Naonori Ueda, "Modeling Network Growth with Directional Attachment and Communities (in Japanese), "IEICE Trans., Vol.J86-D-II, No.10, pp.1468--1479, 2003.
  4. Masahiro Kimura, Kazumi Saito, and Naonori Ueda, "Modeling of growing networks with directional attachment and Communities, "Neural Networks(in press).
  5. Masahiro Kimura, Kazumi Saito, and Naonori Ueda, "Modeling network growth with directional attachment and Communities, "System and Computers in Japan(in press).
  6. Takeshi Yamada, Kazumi Saito, and Naonori Ueda, "Embedding Network Data based on Cross-Entropy Minimization. (in Japanse)," Trans. of IPSJ, Vol.44, No.9, pp.2401--2408, 2003.
  7. Kazumi Saito and Ryohei Nakano, "Squared Penalty Consistend with Linear Transformations of Variables (in Japanse)," Trans. of IPSJ, Vol.44, No.10, pp.2495--2502, 2003.
  8. Pat Langley, Jeff Shrager, and Kazumi Saito, "Computational Discovery of Communicable Scientific Knowledge, "Kluwer Academic,2002.
  9. Kazumi Saito and Ryohei Nakano, "Extracting Regression Rules From Neural Networks (in Japanese), "Neural Networks, Vol.15, pp.1279--1288,2002.
  10. Kazumi Saito, "Data mining using neural networks(in Japanse), "JSIAM,Vol.12, No.4, pp.45--53, 2002.
  11. Ryohei Nakano and Kazumi Saito, "Discovering Polynomials to Fit multivariate Data Having Numeric and Numinal Variables,"Progress in Discovery Science, Vol.LNAI2281, pp.482--493, 2002.
  12. Ryohei Nakano and Kazumi Saito, "Discovering Polynomials to Fit multivariate Data Having Numeric and Numinal Variables (in Japanese), " IEICE Trans., Vol.J84-D-II, No.1, pp.178--187, 2001.
  13. Kazumi Saito and Ryohei Nakano, "Second-order learning algorithm with squared penalty term," Neural Computation, Vol.12, No.3, pp.709--729, 2000.
  14. Kazumi Saito and Ryohei Nakano, "Learning of recurrent networks and estimation of Gaussian mixtures using second-order learning algorithm BPQ (in Japanese)," IEICE Trans., Vol.J81-D-II, No.3, pp.538--546, 1998.
  15. Kazumi Saito and Ryohei Nakano, "A constructive learning algorithm for HME (in Japanese)," IEICE Trans., Vol.J81-D-II, No.2, pp.412--420, 1998.
  16. Kazumi Saito and Ryohei Nakano, "Applying second-order learning algorithm BPQ to classification problems and evaluating it (in Japanese)," IEICE Trans., Vol.J81-D-II, No.2, pp.404--411, 1998.
  17. Kazumi Saito and Ryohei Nakano, "A new regularization based on the MDL principle (in Japanese)," Trans. of JSAI, Vol.13, No.1, pp.123--130, 1998.
  18. Kazumi Saito and Ryohei Nakano, "Second-order learning algorithm with squared penalty term (in Japanese)," Trans. of IPSJ, Vol.38, No.11, pp.2149--2156, pp.1708--1716, 1997.
  19. Kazumi Saito and Ryohei Nakano, "Partial BFGS update and calculating optimal step-length for three-layer neural networks," Neural Computation, Vol.9, No.1, pp.123--141, 1997.
  20. Kazumi Saito and Ryohei Nakano, "A connectionist approach to numeric law discovery (in Japanese)," Trans. of IPSJ, Vol.37, No.9, pp.1708--1716, 1996.
  21. Kazumi Saito and Ryohei Nakano, "Adaptive concept learning algorithm: RF4 (in Japanese)," Trans. of IPSJ, Vol.36, No.4, pp.832--839, 1995.
  22. Kazumi Saito and Ryohei Nakano, "Task sequencing based on Bayesian estimation (in Japanese)," Trans. of IPSJ, Vol.36, No.3, pp.572--578, 1995.
  23. Kazumi Saito and Ryohei Nakano, "Rule extraction from noisy facts: RF3 algorithm (in Japanese)," Trans. of IPSJ, Vol.33, No.5, pp.636--644, 1992.
  24. Kazumi Saito and Ryohei Nakano, "Rule extraction from facts: RF2 algorithm (in Japanse)," Trans. of IPSJ, Vol.33, No.5, pp.628--635, 1992.
  25. Ryohei Nakano and Kazumi Saito, "Reduction of relational calculus to optimal algebraic expression (in Japanse)," Trans. of IPSJ, Vol.29, No.4, pp.397--406, 1988.
  26. Ryohei Nakano and Kazumi Saito, "Reduction of aggregate functions in relational calculus alpha to optimal algebraic expressions (in Japanese)," Trans. of IPSJ, Vol.28, No.12, pp.1246--1254, 1987.

Refereed International Conference Papers

  1. Naonori Ueda and Kazumi Saito. "Simplex Mixture Models for Multi-topic Text," Proc. of The 30th Anniversary of the Information Criterion (AIC 2003) , pp.380--381, 2003.
  2. Masahiro Kimura, Kazumi Saito, and Naonori Ueda. "Modeling of Growing Networks with Directional Attachment and Communities," Proc. of European Symposium on Artificial Neural Networks (ESANN 2003), pp.15--20, 2003.
  3. Takeshi Yamada, Kazumi Saito, and Naonori Ueda. "Cross-Entropy Directed Embedding of Network Data," Proc. of International Conference on Machine Learning (ICML 2003), pp.832--839, 2003.
  4. Masahiro Kimura, Kazumi Saito, and Naonori Ueda. "Modeling Share Dynamics by Extracting Competition Structure," Proc. of the 30th Anniversary of the Information Criterion (AIC 2003) , pp.366--367, 2003.
  5. Pat Langley, Dileep George, Stephen Bay, and Kazumi Saito. "Robust Induction of Process Models from Time-Series Data," Proc. of International Conference on Machine Learning (ICML 2003), pp.432--439, 2003.
  6. Dileep George, Kazumi Satio, Pat Langley, Stephen Bay, and Kevin R. Arrigo. "Discovering Ecosystem Models from Time-Series Data," Proc. of the 6th Discovery Science (DS 2003), pp.141--152, 2003.
  7. Kazumi Saito, Dileep George, Stephen Bay, and Jeff Shrager. "Inducing Biological Models from Temporal Gene Expression Data," Proc. of the 6th Discovery Science (DS 2003), pp.468--469, 2003.
  8. Naonori Ueda and Kazumi Saito, "Parametric Mixture Models for Multi-labeled text," Proc. of MIT press( NIPS)( in press )
  9. Kazumi Saito and Pat Langley, "Qualitative revision of scientific model," Proc. of Computational Discovery of Communicable Knowledge( in press )
  10. Kazumi Saito and Ryohei Nakano, "Structuring Neural Networks through Bidirectional Clustering of Weights," Proc. of the 5th Discovery Science( DS2002 ), pp.206--219, 2002.
  11. Kazumi Saito, Stephen Bay, and Pat Langley, "Revising Qualitative Models of Gene Regulation," Proc. of the 5th Discovery Science( DS2002 ), pp.59--70, 2002.
  12. Masahiro Kimura, Kazumi Saito, and Naonori Ueda, "Modeling of Growing Networks with Communities," Proc. of IEEE( NNSP2002 ), pp.189--198, 2002.
  13. Naonori Ueda and Kazumi Saito, "Single-shot Detection of Multi-category of Text using parametric Mixture Models," Proc. of ACM Special Interest Group( SIGKDD2002 ), pp.626--631, 2002.
  14. Kazumi Saito and Pat Langley, "Discovering Empirical Laws of Web Dynamics," Proc. of IPSJ, IEEE-CS( SAINT-2002 ), pp.168--175, 2002.
  15. Ryohei Nakano and Kazumi Saito, "Finding Polynomials to Fit Multivariate Data Having Numeric and Nominal Variable," Proc. of Advances in Intelligent Data Analysis(IDA2001), Vol.LNCS2189, pp.258--267, 2001.
  16. Kazumi Saito, Pat Langley, Trond Grenager, and Christpher Potter, "Computational Revision of Quantitative scientific models," Proc. of the 3rd International Conference on Discovery Science(DS2001), Vol.LNAI2226, pp.336--349, 2001.
  17. Kazumi Saito and Ryohei Nakano, "178Discovery of a set of nominally conditioned polynomials using neural networks, vector quantizers, and decision trees," Proc. of the 3rd International Conference on Discovery Science (DS00), pp.325--329, 2000. (PS file) (PDF file)
  18. Kazumi Saito and Ryohei Nakano, "Discovery of relevant weights by minimizing cross-validation error," Proc. of the 4th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD2000), pp.372--375, 2000. (PS file) (PDF file)
  19. K.Saito, N.Ueda, S.Katagiri, Y.Fukai, H.Fujimaru, and M.Fujinawa, "Law discovery from financial data using neural networks," Proc of IEEE/IAFE/INFORMS Conference on Computational Intelligence for Financial Engineering (CIFEr00), pp.209--212, 2000 (PS file) (PDF file)
  20. Ryohei Nakano and Kazumi Saito, "Discovery of a set of nominally conditioned polynomials," Proc. of the 2nd International Conference on Discovery Science (DS99), pp.287--298, 1999.
  21. Ryohei Nakano and Kazumi Saito, "Computational characteristics of law discovery using neural networks," Proc. of the 1st International Conference on Discovery Science (DS98), pp.342--351, 1998.
  22. Kazumi Saito and Ryohei Nakano, "Numeric law discovery using neural networks," Proc. of the 4th International Conference on Neural Information Processing (ICONIP97), pp.843--846, 1997. (PS file) (PDF file)
  23. Kazumi Saito and Ryohei Nakano, "Law discovery using neural networks," Proc. of the 15th International Joint Conference on Artificial Intelligence (IJCAI97), pp.1078--1083, 1997. (PS file) (PDF file)
  24. Kazumi Saito and Ryohei Nakano, "MDL regularizer: a new regularizer based on MDL principle," Proc. of International Conference on Neural Networks (ICNN97), pp.1833--1838, 1997. (PS file) (PDF file)
  25. Kazumi Saito and Ryohei Nakano, "Law discovery using neural networks," Proc. of NIPS '96 Rule-Extraction Workshop, pp.62--69, 1996.
  26. Kazumi Saito and Ryohei Nakano, "Second-order learning algorithm with squared penalty term," Proc. of Neural Information Processing Systems Conference (NIPS96), pp.627--633, 1996. (PS file) (PDF file)
  27. Kazumi Saito and Ryohei Nakano, "A constructive learning algorithm for an HME," Proc. of International Conference on Neural Networks (ICNN96), pp.1268--1273, 1996. (PS file) (PDF file)
  28. Ryohei Nakano, Naonori Ueda, Kazumi Saito and M. Takahashi, "Wall map building from fragmentary sonar data," Proc. of RoboLearn '96 Workshop, pp.84--89, 1996.
  29. Ryohei Nakano, Naonori Ueda, Kazumi Saito and Takeshi Yamada, "Parrot-like speaking using optimal vector quantization," Proc. of International Conference on Neural Networks (ICNN95), pp.2871--2875, 1995.
  30. Kazumi Saito and Ryohei Nakano, "Adaptive concept learning algorithm," IFIP Transactions A-Computer Sience and Technology51 (IFIP94). pp.294-299.
  31. Ryohei Nakano, Kazumi Saito, Masakatsu Ohta and S.I. Gallant, "DREAM: A heuristic approach to hypersphere minimum covering," Proc. of International Conference on Artificial Neural Networks (ICANN91), pp.427--432, 1991.
  32. Kazumi Saito and Ryohei Nakano, "Rule extraction from facts and neural nets," Proc. of International Neural Networks Conference (INNC90), pp.379--382, 1990.
  33. Kazumi Saito and Ryohei Nakano, "Medical diagnostic expert system based on PDP model," Proc. of International Conference on Neural Networks (ICNN88), pp.255-262, 1988.

Invited Book Chapters and Articles

  1. Kazumi Saito and Pat Langley, "Quantitative revision of scientific models.," In S. Dzeroski, L. Todorovski,eds.,Computational discovery of communicable scientific knowledge(in Press)
  2. Pat Langley,P.,Shrager,J., Kazumi Saito, " Computational discovery of communicable scientific knowledge," In L.Magnani, N.J.Nersessian, and C.Pizzi, eds., Logical and computational aspects of model-based reasoning,2002.
  3. Kazumi Saito and Ryohei Nakano, "A connectionist approach to numeric law discovery," In K.Furukawa, D.Michie, and S.Muggleton, eds., Machine Intelligence 15 Intelligent Agents (MI15), pp.315--327, Oxford Press, 1999.
  4. Kazumi Saito and Ryohei Nakano, "A concept learning algorithm with adaptive search," In K.Furukawa, D.Michie, and S.Muggleton, eds., Machine Intelligence 14 Applied Machine Intelligence (MI14), pp.347--363, Oxford Press, 1993.
Last modified on: Wed Feb. 19 2003.