Presentation 1997/12/12
Geometrical Analysis of Neural Networks with Asymmetric Weight
Hideki Kakeya, Yoichi Okabe,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In the present paper the geometrical method based on the eigenspace analysis is applied to analyze the dynamics of neural networks with asymmetric weight, matrices. Crosscorrelational associative memory and random networks are investigated and complex dynamical behavior is explained from the geometrical viewpoint. Nonmonotonic dynamics for asymmetric networks are proposed based on the geometrical discussion.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) neuro-dynamics / asymmetric matrix / eigenspace analysis / crosscorrelational associative memory / random network / nonmonotonic dynamics
Paper # NC97-57
Date of Issue

Conference Information
Committee NC
Conference Date 1997/12/12(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Geometrical Analysis of Neural Networks with Asymmetric Weight
Sub Title (in English)
Keyword(1) neuro-dynamics
Keyword(2) asymmetric matrix
Keyword(3) eigenspace analysis
Keyword(4) crosscorrelational associative memory
Keyword(5) random network
Keyword(6) nonmonotonic dynamics
1st Author's Name Hideki Kakeya
1st Author's Affiliation Research Center for Advanced Science and Technology, University of Tokyo()
2nd Author's Name Yoichi Okabe
2nd Author's Affiliation Research Center for Advanced Science and Technology, University of Tokyo
Date 1997/12/12
Paper # NC97-57
Volume (vol) vol.97
Number (no) 448
Page pp.pp.-
#Pages 8
Date of Issue