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 2012-03-14 13:20
Training Multi-layered Neural Network Neocognitron
Kunihiko Fukushima NC2011-128
Abstract (in Japanese) (See Japanese page) 
(in English) This paper proposes new learning rules suited for training multi-layered neural networks and apply them to the neocognitron. The neocognitron is a hierarchical multi-layered neural network capable of robust visual pattern recognition. It acquires the ability to recognize visual patterns through learning. For training intermediate stages of the hierarchical network of the neocognitron, we use a new learning rule named add-if-silent. For the highest stage of the network, we use the method of \textit{interpolating vectors}, not only during recognition, but also during learning. By the use of add-if-silent rule, the learning algorithm is considerably simplified, and parameters required in designing the network can be reduced. The method of interpolating vectors greatly increases the recognition rate. Computer simulation demonstrates that the new neocognitron shows a higher recognition rate with a smaller scale of the network than the neocognitron of previous versions.
Keyword (in Japanese) (See Japanese page) 
(in English) visual pattern recognition / neocognitron / hierarchical neural network / learning rule / add-if-silent / interpolating vectors / /  
Reference Info. IEICE Tech. Rep., vol. 111, no. 483, NC2011-128, pp. 39-44, March 2012.
Paper # NC2011-128 
Date of Issue 2012-03-07 (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 NC2011-128

Conference Information
Committee MBE NC  
Conference Date 2012-03-14 - 2012-03-16 
Place (in Japanese) (See Japanese page) 
Place (in English) Tamagawa University 
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To NC 
Conference Code 2012-03-MBE-NC 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Training Multi-layered Neural Network Neocognitron 
Sub Title (in English)  
Keyword(1) visual pattern recognition  
Keyword(2) neocognitron  
Keyword(3) hierarchical neural network  
Keyword(4) learning rule  
Keyword(5) add-if-silent  
Keyword(6) interpolating vectors  
Keyword(7)  
Keyword(8)  
1st Author's Name Kunihiko Fukushima  
1st Author's Affiliation * (*)
2nd Author's Name  
2nd Author's Affiliation ()
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 2012-03-14 13:20:00 
Presentation Time 25 minutes 
Registration for NC 
Paper # NC2011-128 
Volume (vol) vol.111 
Number (no) no.483 
Page pp.39-44 
#Pages
Date of Issue 2012-03-07 (NC) 


[Return to Top Page]

[Return to IEICE Web Page]


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