Presentation 2002/1/22
Probabilistic Information Processing and Statistical Physics
Jun-ichi INOUE,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Nowadays we all know the importance of massive information processing, including information and communication, image processing, optimization etc.. As a candidate of the effective tools to treat these problems, statistical mechanics has attracted a great deal of attention. Statistical mechanical technique enable us to construct some probabilistic algorithms to solve the problems which are represented by probabilistic models. In addition, by using statistical mechanics, we can evaluate the performance of the algorithms extensively. In this talk, I show the relation between the probabilistic information processing and statistical physics in terms of Bayesian statistics, and explain what can be done by statistical physics from three different view points, namely, probabilistic modeling of the system, constructing mean-field algorithm and evaluating the statistical performance by so-called replica method.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Statistical mechanics / Probabilistic model / Bayesian statistics / Mean-field approximation / Replica method
Paper #
Date of Issue

Conference Information
Committee NC
Conference Date 2002/1/22(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 Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Probabilistic Information Processing and Statistical Physics
Sub Title (in English)
Keyword(1) Statistical mechanics
Keyword(2) Probabilistic model
Keyword(3) Bayesian statistics
Keyword(4) Mean-field approximation
Keyword(5) Replica method
1st Author's Name Jun-ichi INOUE
1st Author's Affiliation Graduate School of Engineering, Hokkaido University()
Date 2002/1/22
Paper #
Volume (vol) vol.101
Number (no) 616
Page pp.pp.-
#Pages 6
Date of Issue