Effect of time correlation of input patterns on the convergence of on-line learning

Physical Review E - Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics

Volumn 53, Issue 4, 1996, Pages 4217-4220

  Hondou, Tsuyoshi ^a   Yamamoto, Mitsuaki ^a   Sawada, Yasuji ^a   Hayakawa, Yoshihiro ^a  

^a TOHOKU UNIVERSITY   (Japan)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0040389085     PISSN: 1063651X     EISSN: None     Source Type: Journal    
DOI: 10.1103/PhysRevE.53.4217     Document Type: Article

References (20)

1. G.J. Mpitsos and R.M. Burton, Neural Networks5, 605 (1992).
2. T. Hondou and Y. Sawada, Prog. Theor. Phys.91, 397 (1994).
3. H. Nakajima and Y. Ueda, in Proceedings of the International Conference on Nonlinear Theory and Its Application (Institute of Electronics, Information and Communication Engineers, Tokyo, 1993), p. 601.
4. D.E. Rumelhart, J.L. McClelland, and the PDP Research Group, Parallel Distributed Processing (MIT Press, Cambridge, 1986).
5. G. Radons, H.G. Schuster, and D.Werner, in Parallel Processing in Neural Systems and Computers, edited by R. Eckmiller (Elsevier, Amsterdam, 1990), p. 261.
6. T.M. Heskes and B. Kappen, Phys. Rev. A44, 2718 (1991).
7. L.K. Hansen, R. Pathria and P. Salamon, J. Phys. A26, 63 (1993).
8. G. Radons, J. Phys. A26, 3455 (1993).
9. M. Biehl and H. Schwarze, J. Phys. A28, 643 (1995).
10. D. Saad and S.A. Solla, Phys. Rev. Lett.74, 4337 (1995).
11. W. Wiegerinck and T. Heskes, Europhys. Lett.28, 451 (1994).
12. See, for example, H.G. Schuster, Deterministic Chaos (Physik-Verlag, Weinheim, Germany, 1984).
13. Y. Hayakawa, A. Marumoto and Y. Sawada, Phys. Rev. E51, 2693 (1995).
14. T. Hondou, J. Phys. Soc. Jpn.63, 2014 (1994);
15. T. Hondou and Y. Sawada, Phys. Rev. Lett.75, 3269 (1995);
16. T. Hondou and Y. Sawada, Phys. Rev. Lett.76, 1005 (1996).
17. S. Amari, IEEE Trans.EC-16, 299 (1967).
18. H.J. Kushner, SIAM. J. Appl. Math.44, 160 (1984).
19. In feedforward neural networks with no less than three layers, any continuous function can be approximated with any precision. Proof appears in K. Funahashi, Neural Networks2, 183 (1989).
20. An example is found in Fig. 4 of Ref. 2; T. Hondou, Prog. Theor. Phys. (to be published).