Haruhisa Takahashi

Professor
Department of Information and Communication Engineering
University of Electro-Communications
1-5-1 Chofugaoka, Chofushi, Tokyo 182, Japan
e-mail:takahasi@sw.cas.uec.ac.jp
Tel: +81-424-43-5244;
Fax: +81-424-89-6943
Office: W8 room 807

Biography

He was born in Shizuoka Prefecture Japan, on March 31, 1952. He graduated from the University of Electro-Communications, Tokyo. He received the Dr Eng. degree from Osaka University. He was a faculty member of the Department of Computer Science and Engineering, Toyohashi University of Technology from 1980 to 1986. Since 1986, he has been with the University of Electro-Communications, where he is currently Professor of Communications and Systems Engineering. He was previously engaged in the field of nonlinear network theory, queueing theory and performance evaluation of communication systems.

Current research interests

• learning machines and neural networks (mathematical theory and applications to speech recognition and image processing)
• cognitive sciences

Journal papers and proceedings

• Recent journal papers in Neural Networks( Only for english publications)
  • H. Takahashi and E. Tomita, T. Kawabata: Separability of Internal Representations in Multilayer Perceptrons with Application to Learning. Neural Networks 6 , 689-703 (1993).
  • H. Gu , H. Takahashi : Towards More Practical Average Bounds on Supervised Learning. IEEE Trans. Neural Networks 7, No.5, 953-968 (1996)
  • H . Gu, H. Takahashi: Learning Curves in Learning with Noise - An Empirical Study, Trans. Inst. Electron. Inf. and Commun. Engrs. Japan, Informatin and Systems, Vol. E-80-D, No.1, 78-85 (1997).
  • H. Gu, H. Takahashi: Estimating Learning Curves of Concept Learning, Neural Networks 10, No.6, 1089-1102 (1997)
  • H. Takahashi: A Tighter Upper Bound on Storage Capacity of Multilayer Neural Networks, Trans. Inst. Electron. Inf. and Commun. Engrs. Japan, Fundamentals Vol E81-A. No.2, 333-339 (1998).
  • H. Takahashi and H. Gu: A Tight Bound on Concept Learning, IEEE Trans. Neural Networks Vol 9, 6, Nov. , 1191-1202 (1998)
  • H. Gu and H. Takahashi: Exponential or Polynomial Learning Curves? Case-Based Studies, Neural Computation (to appear in Neural Computation).

• (manuscripts)
  • PAC Learning via Ill-Disposed Algorithm, Haruhisa Takahashi and Hanzhong Gu, submitted to Machine Learning. (ps file is available)
  • (besides 8 international proceedings and 19 technical reports of IEICE)

• Algorithm
  • Tomita, E, Takahashi H, Mitsuma, S A polynomial time algorithm finding near maximum clique, IEICE J74-D-I, No.4 (1991): 307-310. (in Japanese)
  • Yamada Y, Tomita, E, Takahashi H, "A randomized algorithm for finding a near-maximum clique and its experimental evaluations", IEICE, D-I, Vol J76-D-I, No.2 (1993). (in Japanese)
  • (besides other related 4 proceedings)
• nonlinear networks
  • Takahashi, H, Kumagai S, Kodama, S "Generalized decomposition theorem of nonlinear n-port networks", Trans. Inst. Electron. Commun. Engrs. Japan, 61-A8｡｢698-705,｡｡｡ﾊ1978) (in Japanese). (besides 6 other related papers and 3 proceedings)
• Queueing theory and Performance evaluation of communication networks
  • H. Takahashi, Y. Takahashi: Notes on Conservation Laws for Preemptive Priority Queues, Trans. Inst. Electron. Inf. and Commun. Engrs. Japan, E72-8, 891-897, (1989).
  • H. Takahashi: Coupled Processor - A Second Order Continuous State Space Model, Probability in the Eng. and Inf. Sciences No.2, 277-298, (1990).
  • (besides other related 16 papers and 22 proceedings)
• Translation
  • Analysis and Synthesis of Computer Systems, Academic Press (London), 1980, Gelenbe, Mitrani. (from English to Japanese)
• Proceedings
  • H. Takahashi and E. Tomita: Estimation of Learning Curve in Learning Neural Networks From Noisy Sample. International Symposium on Nonlinear Theory and its Applications, HAWAII , 47-50 (1993).
  • H. Takahashi and E. Tomita: Estimating Learning Curves by PAC-Learnability Criterion. International Joint Conference on Neural Networks, Nagoya Vol.2 , Oct. 1641-1644 (1993)
  • H. Takahashi: Statistical -Capacity of Multilayer Networks and Generalization Performance. Proc. Symposium on Nonlinear Theoory and Its Applications '94, 271-274 (1994).
  • H. Gu and H. Takahashi: Generalizations of the PAC Model for Function Estimation, Proc. Symposium on Nonlinear Theoory and Its Applications '94, 267-270 (1994)
  • H. Takahashi: Towards Practical Bounds of Storage Capacity in Multilayer Networks, Proceedings of 1995 Simposium on Nonlinear Theory and its Applications (NOLTA'95), Caesars Palace, Las Vegas, USA, Dec., 215-218 (1995).
  • H. Gu and H. Takahashi: Exponential or Polynomial Learning Curves? A Case Study, Proceedings of 1995 Simposium on Nonlinear Theory and its Applications, Caesars Palace, Las Vegas, USA, Dec., 243-246 (1995).
  • I. V. Mayer, H. Takahashi: Color Edge Detection via Neural Networks, Progress in Connectionist-Based Information Systems, Vol.2 N.Kasabov, et al ed., Springer, 1195-1198 (1997).
  • H. Takahashi, T. Uchiyama: Biologically Plausible and Efficient Learning, Progress in Connectionist-Based Information Systems, Vol.1 N.Kasabov, et al ed., Springer, 259-262 (1997).
  • I., V. Mayer, H. Takahashi: Color Edge Detection via Neural Networks, Proc. International Symposium on Nonlinear Theoory and Its Applications '97, Hawaii, 1021-1024 (1997).
  • H. Takahashi, T. Uchiyama: Biologically Plausible Local Error Correction Learning, Proc. International Symposium on Nonlinear Theoory and Its Applications '97, Hawaii, 723-726 (1997).
  • I., V. Mayer, H. Takahashi: Matching by Principal Component Analysis, The Second International Conference on Cognitive and Neural Systems, Boston, May (abstract) (1998).

  ---

  

  Comments or suggestions would highly be appreciated. Last updated: June 9, 1999.
  Back to "Laboratory members"