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 2012-11-17 13:30
Composite likelihood estimation for bipartite Boltzmann machines
Takashi Asari, Muneki Yasuda, Yuji Waizumi, Kazuyuki Tanaka (Tohoku Univ.) NC2012-68
Abstract (in Japanese) (See Japanese page) 
(in English) The recent development of information technology has enabled us to obtain and to storage huge information data.
Because of this, effective usages of the information data have been one of the central points of information sciences.
A Boltzmann machine is a model of the machine learning theory and can be a powerful tool for the aim.
In this paper, we focus on a bipartite Boltzmann machine (BBM) consisted of two different layers: one layer is the input layer and the other is the output layer.
In this model, once the input layer is clamped by given data, inference of the output layer is quite easy because of the property of conditional independence.
Therefore, this model is expected to be applied to expert systems such as medical test system.
However, the learning of BBM using the maximum likelihood estimation is still intractable,
Because the computational complexity of learning exponentially grows with the increase in the size of system,
In this paper, we propose a new learning algorithm for BBM by using the composite likelihood estimation (CLE) which is a statistical mathematical technique to approximate the MLE.
Keyword (in Japanese) (See Japanese page) 
(in English) probabilistic information processing / statistical machine learning theory / Boltzmann machine / maximum likelihood estimation / composite likelihood estimation / / /  
Reference Info. IEICE Tech. Rep., vol. 112, no. 298, NC2012-68, pp. 39-44, Nov. 2012.
Paper # NC2012-68 
Date of Issue 2012-11-09 (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 NC2012-68

Conference Information
Committee MBE NC  
Conference Date 2012-11-16 - 2012-11-17 
Place (in Japanese) (See Japanese page) 
Place (in English) Tohoku University 
Topics (in Japanese) (See Japanese page) 
Topics (in English) BCI/BMI, NC and MBE 
Paper Information
Registration To NC 
Conference Code 2012-11-MBE-NC 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Composite likelihood estimation for bipartite Boltzmann machines 
Sub Title (in English)  
Keyword(1) probabilistic information processing  
Keyword(2) statistical machine learning theory  
Keyword(3) Boltzmann machine  
Keyword(4) maximum likelihood estimation  
Keyword(5) composite likelihood estimation  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Takashi Asari  
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 Yuji Waizumi  
3rd Author's Affiliation Tohoku University (Tohoku Univ.)
4th Author's Name Kazuyuki Tanaka  
4th Author's Affiliation Tohoku University (Tohoku Univ.)
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 2012-11-17 13:30:00 
Presentation Time 25 minutes 
Registration for NC 
Paper # NC2012-68 
Volume (vol) vol.112 
Number (no) no.298 
Page pp.39-44 
#Pages
Date of Issue 2012-11-09 (NC) 


[Return to Top Page]

[Return to IEICE Web Page]


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