Presentation 1999/9/21
Searching Ability of a Solution of Combinatorial Optimization Problems by Hysteresis Neural Networks.
Kenya Jin'no,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We consider combinational optimization problems solver by using hysteresis neural networks. First, hysteresis neural networks were compared with conventional neural networks. The characteristic of the equilibrium point of these systems is analyzed. And then, hysteresis neural networks can do that the equilibrium point is made to comply only with the optimal solution of a combinatorial optimization problem. Hystsresis neural networks don't guarantee its energy to decrease monotonously. Namely, the output of this system may oscillate. Such oscillational state can be restrained to control the time constant of each hysteresis neuron.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) time constants / neural networks / hysteresis / domain of attraction / combinatorial optimization problems / self connection
Paper # NLP99-84
Date of Issue

Conference Information
Committee NLP
Conference Date 1999/9/21(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 JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Searching Ability of a Solution of Combinatorial Optimization Problems by Hysteresis Neural Networks.
Sub Title (in English)
Keyword(1) time constants
Keyword(2) neural networks
Keyword(3) hysteresis
Keyword(4) domain of attraction
Keyword(5) combinatorial optimization problems
Keyword(6) self connection
1st Author's Name Kenya Jin'no
1st Author's Affiliation Department of Electrical Electronic Engineering, Nippon Institute of Technology()
Date 1999/9/21
Paper # NLP99-84
Volume (vol) vol.99
Number (no) 323
Page pp.pp.-
#Pages 6
Date of Issue