Presentation 1999/7/22
An Analysis of an Associative Memory with Sparse Connection Matrix
Naoki MASUDA, Kazuyuki AIHARA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Associative memories have been studied to large extent: theoretical analysis, extention of the model, analysis in the case of sparse connection, application to pattern analysis. An associative memory with a sparse connection matrix are biologically more plausible than the associative memory connected in the all-to-all manner, and it leads to the reduction of implementation cost. In this report, the dynamics of an associative memory with a sparsely connected matrix is analyzed theoretically. Geometrical analysis and statistical analysis show that less connectivity indicates the unstablity of memorized pattern vectors and lessens the size of an attractive basin.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) associative memory / connection matrix / sparse connection / equilibrium / basin of attractor
Paper # NLP99-56
Date of Issue

Conference Information
Committee NLP
Conference Date 1999/7/22(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 Nonlinear Problems (NLP)
Language ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) An Analysis of an Associative Memory with Sparse Connection Matrix
Sub Title (in English)
Keyword(1) associative memory
Keyword(2) connection matrix
Keyword(3) sparse connection
Keyword(4) equilibrium
Keyword(5) basin of attractor
1st Author's Name Naoki MASUDA
1st Author's Affiliation Department of Mathematical Engineering and Information Physics, Graduate School of Engineering, The University of Tokyo()
2nd Author's Name Kazuyuki AIHARA
2nd Author's Affiliation CREST, Japan Science and Technology Co. (JST)
Date 1999/7/22
Paper # NLP99-56
Volume (vol) vol.99
Number (no) 204
Page pp.pp.-
#Pages 8
Date of Issue