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 2010-11-05 15:30
[Poster Presentation] Statistical Mechanics of Adaptive Weight Perturbation Learning
Ryousuke Miyoshi, Yutaka Maeda, Seiji Miyoshi (Kansai Univ.) IBISML2010-92
Abstract (in Japanese) (See Japanese page) 
(in English) The weight perturbation learning was proposed as a learning rule which adds perturbation to the variable parameters of learning machines. Generalization performance of the weight perturbation learning was analyzed by statistical mechanical methods. The weight perturbation learning has the same asymptotic generalization property as the Perceptron learning. In this paper we consider difference between the Perceptron learning and the AdaTron learning which are well-known learning rules. Applying the consideration to the weight perturbation learning, we propose the adaptive weight perturbation learning. The generalization performance of the proposed rule is analyzed by statistical mechanical methods. Consequently, it is shown that the proposed learning rule has an outstanding asymptotic property corresponding to the AdaTron learning.
Keyword (in Japanese) (See Japanese page) 
(in English) on-line learning / weight perturbation / simultaneous perturbation / generalization error / statistical mechanics / / /  
Reference Info. IEICE Tech. Rep., vol. 110, no. 265, IBISML2010-92, pp. 245-250, Nov. 2010.
Paper # IBISML2010-92 
Date of Issue 2010-10-28 (IBISML) 
ISSN Print edition: ISSN 0913-5685  Online edition: ISSN 2432-6380
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 IBISML2010-92

Conference Information
Committee IBISML  
Conference Date 2010-11-04 - 2010-11-06 
Place (in Japanese) (See Japanese page) 
Place (in English) IIS, Univ. of Tokyo 
Topics (in Japanese) (See Japanese page) 
Topics (in English) IBIS 2010 (Workshop on Information-based Induction Sciences) 
Paper Information
Registration To IBISML 
Conference Code 2010-11-IBISML 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Statistical Mechanics of Adaptive Weight Perturbation Learning 
Sub Title (in English)  
Keyword(1) on-line learning  
Keyword(2) weight perturbation  
Keyword(3) simultaneous perturbation  
Keyword(4) generalization error  
Keyword(5) statistical mechanics  
1st Author's Name Ryousuke Miyoshi  
1st Author's Affiliation Kansai University (Kansai Univ.)
2nd Author's Name Yutaka Maeda  
2nd Author's Affiliation Kansai University (Kansai Univ.)
3rd Author's Name Seiji Miyoshi  
3rd Author's Affiliation Kansai University (Kansai 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 ()
Date Time 2010-11-05 15:30:00 
Presentation Time 180 
Registration for IBISML 
Paper # IEICE-IBISML2010-92 
Volume (vol) IEICE-110 
Number (no) no.265 
Page pp.245-250 
#Pages IEICE-6 
Date of Issue IEICE-IBISML-2010-10-28 

[Return to Top Page]

[Return to IEICE Web Page]

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