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 2014-11-17 17:00
[Poster Presentation] Feature Extraction for Image Classification using Restricted Boltzmann Machines
Reiki Suda, Koujin Takeda (Ibaraki Univ.) IBISML2014-36
Abstract (in Japanese) (See Japanese page) 
(in English) Learning restricted Boltzmann machines (RBMs) for high-dimensional data using maximum likelihood estimation had been faced with computational complexity.
However, after the development of approximation algorithms based on Markov chain Monte Carlo methods, RBMs have been applicable even to high-dimensional data, and playing a central role as feature extractors.
In this study, we conduct experiments of feature extraction from handwritten digit images using RBMs and of classification with the extracted features.
As classifiers, fast and simple linear classifiers based on confidence-weighted learning are used.
Finally, we show that classifiers trained with features extracted by RBMs outperform ones trained with raw data, and from the result we also evaluate the learning performance of the RBMs.
Keyword (in Japanese) (See Japanese page) 
(in English) restricted Boltzmann machines / image recognition / feature extraction / Markov chain Monte Carlo methods / linear classifiers / confidence-weighted learning / /  
Reference Info. IEICE Tech. Rep., vol. 114, no. 306, IBISML2014-36, pp. 9-15, Nov. 2014.
Paper # IBISML2014-36 
Date of Issue 2014-11-10 (IBISML) 
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 IBISML2014-36

Conference Information
Committee IBISML  
Conference Date 2014-11-17 - 2014-11-19 
Place (in Japanese) (See Japanese page) 
Place (in English) Nagoya Univ. 
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To IBISML 
Conference Code 2014-11-IBISML 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Feature Extraction for Image Classification using Restricted Boltzmann Machines 
Sub Title (in English)  
Keyword(1) restricted Boltzmann machines  
Keyword(2) image recognition  
Keyword(3) feature extraction  
Keyword(4) Markov chain Monte Carlo methods  
Keyword(5) linear classifiers  
Keyword(6) confidence-weighted learning  
Keyword(7)  
Keyword(8)  
1st Author's Name Reiki Suda  
1st Author's Affiliation Ibaraki University (Ibaraki Univ.)
2nd Author's Name Koujin Takeda  
2nd Author's Affiliation Ibaraki University (Ibaraki Univ.)
3rd Author's Name  
3rd Author's Affiliation ()
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 2014-11-17 17:00:00 
Presentation Time 180 minutes 
Registration for IBISML 
Paper # IBISML2014-36 
Volume (vol) vol.114 
Number (no) no.306 
Page pp.9-15 
#Pages
Date of Issue 2014-11-10 (IBISML) 


[Return to Top Page]

[Return to IEICE Web Page]


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