Presentation 2008-06-27
A Visualized Analysis on State Transition in Hopfield Network : Search Process of Combinatorial Optimization Problems
Kiyotaka YAMASHITA, Hiroshi WAKUYA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A Hopfield network is a good tool for solving combinatorial optimization problems, even though it sometimes fails to find out an optimal solution. According to the preceding study, a virtual magnetic diminuendo method proposed recently improves the score through controlling a threshold of the network as a magnetic parameter. Then, in order to make clear its mechanism, behavior of the Hopfield network is analyzed carefully in this article. As a result, it is found that i) a critical magnetic parameter separates firing and quiescent regions, ii) crossing this critical point makes the model control the easiness of neurons' activity, and iii) it leads the model easier to reach the goal.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Hopfield network / combinatorial optimization problem / virtual magnetic diminuendo method / magnetic parameter / critical point
Paper # NC2008-28
Date of Issue

Conference Information
Committee NC
Conference Date 2008/6/19(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) A Visualized Analysis on State Transition in Hopfield Network : Search Process of Combinatorial Optimization Problems
Sub Title (in English)
Keyword(1) Hopfield network
Keyword(2) combinatorial optimization problem
Keyword(3) virtual magnetic diminuendo method
Keyword(4) magnetic parameter
Keyword(5) critical point
1st Author's Name Kiyotaka YAMASHITA
1st Author's Affiliation Faculty of Science and Engineering, Saga University()
2nd Author's Name Hiroshi WAKUYA
2nd Author's Affiliation Faculty of Science and Engineering, Saga University
Date 2008-06-27
Paper # NC2008-28
Volume (vol) vol.108
Number (no) 101
Page pp.pp.-
#Pages 6
Date of Issue