Presentation 2007-03-14
Analysis of Exchange Ratio for Heat Bath Type in Exchange Monte Carlo Method
Kenji NAGATA, Sumio WATANABE,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The exchange Monte Carlo method is well known as an improved algorithm of Markov Chain Monte Carlo method. Although its effectiveness has been shown in many fields, the mathmatical foundation of exchange Monte Carlo method has not yet been established. In our previous work, we analytically clarify the asymptotic behavior of symmetirized Kullback divergence and exchange ratio for Metropolis type in low temperature limit. In this paper, we analytically charify the asymptotic behavior of the exchange ratio for heat bath type in low temperature limit, and discuss the mathematical property of exchange ratio for Metropolis type and for heat bath type.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Markov Chain Monte Carlo Method / Exchange Monte Carlo Method / Kullback Divergence / Exchange Ratio / Average Exchange Ratio
Paper # NC2006-144
Date of Issue

Conference Information
Committee NC
Conference Date 2007/3/7(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 ENG
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Analysis of Exchange Ratio for Heat Bath Type in Exchange Monte Carlo Method
Sub Title (in English)
Keyword(1) Markov Chain Monte Carlo Method
Keyword(2) Exchange Monte Carlo Method
Keyword(3) Kullback Divergence
Keyword(4) Exchange Ratio
Keyword(5) Average Exchange Ratio
1st Author's Name Kenji NAGATA
1st Author's Affiliation Department of Computer Science Tokyo Institute of Technology()
2nd Author's Name Sumio WATANABE
2nd Author's Affiliation PI Lab., Tokyo Institute of Technology
Date 2007-03-14
Paper # NC2006-144
Volume (vol) vol.106
Number (no) 588
Page pp.pp.-
#Pages 6
Date of Issue