Presentation 2002/5/13
Statistical Property of a Boltzmann Machine with Nonmonotonic Neurons
Takuya HAGA, Fumihiko ISHIDA, Mitsunaga KINJO, Shigeo SATO, Koji NAKAJIMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) It is known that nonmonotonic neurons improve their learning performance in a deterministic Boltzmann machine. But there have not been enough theoretical analyses about the optimum learning algorithm and threshold for nonmonotonic neurons. To study them, we investigate a property of statistical mechanics of the Boltzmann machine with nonmonotonic neurons, because the learning algorithm is derived from statistical property. Computer simulations show that the energy distribution at equilibrium states is given by the Boltzmann distribution also with nonmonotonic neurons. From these results, we show a possibility of obtaining the optimum threshold for nonmonotonic neurons.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Neural network / Boltzmann machines / Nonmonotonic neuron / Learning / Boltzmann distribution
Paper # NLP2002-2
Date of Issue

Conference Information
Committee NLP
Conference Date 2002/5/13(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) Statistical Property of a Boltzmann Machine with Nonmonotonic Neurons
Sub Title (in English)
Keyword(1) Neural network
Keyword(2) Boltzmann machines
Keyword(3) Nonmonotonic neuron
Keyword(4) Learning
Keyword(5) Boltzmann distribution
1st Author's Name Takuya HAGA
1st Author's Affiliation Research Institute of Electrical Communication, Tohoku University Laboratory for Electronic Intelligent Systems, Research Institute of Electrical Communication, Tohoku University()
2nd Author's Name Fumihiko ISHIDA
2nd Author's Affiliation Research Institute of Electrical Communication, Tohoku University Laboratory for Electronic Intelligent Systems, Research Institute of Electrical Communication, Tohoku University
3rd Author's Name Mitsunaga KINJO
3rd Author's Affiliation Research Institute of Electrical Communication, Tohoku University Laboratory for Electronic Intelligent Systems, Research Institute of Electrical Communication, Tohoku University
4th Author's Name Shigeo SATO
4th Author's Affiliation Research Institute of Electrical Communication, Tohoku University Laboratory for Electronic Intelligent Systems, Research Institute of Electrical Communication, Tohoku University
5th Author's Name Koji NAKAJIMA
5th Author's Affiliation Research Institute of Electrical Communication, Tohoku University Laboratory for Electronic Intelligent Systems, Research Institute of Electrical Communication, Tohoku University
Date 2002/5/13
Paper # NLP2002-2
Volume (vol) vol.102
Number (no) 68
Page pp.pp.-
#Pages 6
Date of Issue