Presentation 2002/3/12
The Method to solve Optimization Problems on a Neural Network using Inverse Delayed Model
Yoshihiro HAYAKAWA, Koji NAKAJIMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) It is an elementary assumption to have two states resting and excitation for neurons in the past network models. And it seems few discussions have been presented what rolls the neuron in the critical state has on the network actions. Therefore, we have presented the novel model (Inverse Delayed Model) to represent neurons in the critical state, and discuss the fundamental characteristics of the ID model. In this paper, we discuss to use the characteristic of the ID model's negative resistance for escaping from local minima of the optimization problem on a neural network. As a result we present the neural network using ID model to be able to answer correct solution in an only fixed state of this network by setting status of local minimum in negative resistance region.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Neural Networks / Negative Resistance / Optimization Problem / BVP model
Paper # NC2001-187
Date of Issue

Conference Information
Committee NC
Conference Date 2002/3/12(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) The Method to solve Optimization Problems on a Neural Network using Inverse Delayed Model
Sub Title (in English)
Keyword(1) Neural Networks
Keyword(2) Negative Resistance
Keyword(3) Optimization Problem
Keyword(4) BVP model
1st Author's Name Yoshihiro HAYAKAWA
1st Author's Affiliation Laboratory for Electronic Intelligent Systems Research Institute of Electrical Communication, Tohoku University()
2nd Author's Name Koji NAKAJIMA
2nd Author's Affiliation Laboratory for Electronic Intelligent Systems Research Institute of Electrical Communication, Tohoku University
Date 2002/3/12
Paper # NC2001-187
Volume (vol) vol.101
Number (no) 736
Page pp.pp.-
#Pages 8
Date of Issue