IEICE Technical Committee Submission System
Conference Paper's Information
Online Proceedings
[Sign in]
Tech. Rep. Archives
 Go Top Page Go Previous   [Japanese] / [English] 

Paper Abstract and Keywords
Presentation 2009-03-12 14:15
Learning algorithm in Boltzmann machines using the belief propagation algorithm
Junya Tannai, Muneki Yasuda, Kazuyuki Tanaka (Tohoku Univ.) NC2008-144
Abstract (in Japanese) (See Japanese page) 
(in English) Boltzmann machines are stochastic neural networks defined on undirected graphs and are expected to be learning machines with rich structures. It is known that the contrastive divergence learning is one of high-speed and effective learning algorithms for fully-visible Boltzmann machines and restricted Boltzmann machines. However, there are few effective learning algorithms for general Boltzmann machines which have some connections among hidden units. In the present paper, we propose a new learning algorithm for Boltzmann machines with general structures using the contrastive divergence and the loopy belief propagation. In some numerical experiments for artificial data, we show its validity by comparing our proposed algorithm with the mean-field learning one by Welling and Hinton.
Keyword (in Japanese) (See Japanese page) 
(in English) Boltzmann machine / Markov random field / belief propagation / contrastive divergence / / / /  
Reference Info. IEICE Tech. Rep., vol. 108, no. 480, NC2008-144, pp. 243-248, March 2009.
Paper # NC2008-144 
Date of Issue 2009-03-04 (NC) 
ISSN Print edition: ISSN 0913-5685    Online edition: ISSN 2432-6380
Copyright
and
reproduction
All rights are reserved and no part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage and retrieval system, without permission in writing from the publisher. Notwithstanding, instructors are permitted to photocopy isolated articles for noncommercial classroom use without fee. (License No.: 10GA0019/12GB0052/13GB0056/17GB0034/18GB0034)
Download PDF NC2008-144

Conference Information
Committee NC MBE  
Conference Date 2009-03-11 - 2009-03-13 
Place (in Japanese) (See Japanese page) 
Place (in English) Tamagawa Univ. 
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To NC 
Conference Code 2009-03-NC-MBE 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Learning algorithm in Boltzmann machines using the belief propagation algorithm 
Sub Title (in English)  
Keyword(1) Boltzmann machine  
Keyword(2) Markov random field  
Keyword(3) belief propagation  
Keyword(4) contrastive divergence  
Keyword(5)  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Junya Tannai  
1st Author's Affiliation Tohoku University (Tohoku Univ.)
2nd Author's Name Muneki Yasuda  
2nd Author's Affiliation Tohoku University (Tohoku Univ.)
3rd Author's Name Kazuyuki Tanaka  
3rd Author's Affiliation Tohoku University (Tohoku Univ.)
4th Author's Name  
4th Author's Affiliation ()
5th Author's Name  
5th Author's Affiliation ()
6th Author's Name  
6th Author's Affiliation ()
7th Author's Name  
7th Author's Affiliation ()
8th Author's Name  
8th Author's Affiliation ()
9th Author's Name  
9th Author's Affiliation ()
10th Author's Name  
10th Author's Affiliation ()
11th Author's Name  
11th Author's Affiliation ()
12th Author's Name  
12th Author's Affiliation ()
13th Author's Name  
13th Author's Affiliation ()
14th Author's Name  
14th Author's Affiliation ()
15th Author's Name  
15th Author's Affiliation ()
16th Author's Name  
16th Author's Affiliation ()
17th Author's Name  
17th Author's Affiliation ()
18th Author's Name  
18th Author's Affiliation ()
19th Author's Name  
19th Author's Affiliation ()
20th Author's Name  
20th Author's Affiliation ()
Speaker Author-1 
Date Time 2009-03-12 14:15:00 
Presentation Time 25 minutes 
Registration for NC 
Paper # NC2008-144 
Volume (vol) vol.108 
Number (no) no.480 
Page pp.243-248 
#Pages
Date of Issue 2009-03-04 (NC) 


[Return to Top Page]

[Return to IEICE Web Page]


The Institute of Electronics, Information and Communication Engineers (IEICE), Japan