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 2023-03-03 16:25
Fast Identification of Possible Model Parameter Update for Low-Rank Update of Training Data
Hiroyuki Hanada, Noriaki Hashimoto (RIKEN), Kouichi Taji, Ichiro Takeuchi (Nagoya Univ.) PRMU2022-123 IBISML2022-130
Abstract (in Japanese) (See Japanese page) 
(in English) Machine learning methods often require re-training the training dataset with low-rank modifications (small number of instance/feature additions or removals), like leave-one-out cross-validation (LOOCV) and stepwise feature selection. For such modifications, we consider identifying the possible region of updated model parameters instead of calculating them exactly, for fast computation. We first present the method and the condition to identify the region for regularized empirical risk minimization problems, and then how to make LOOCV and stepwise feature selection faster.
Keyword (in Japanese) (See Japanese page) 
(in English) Low-rank update of data / Leave-one-out cross-validation / Stepwise feature selection / Regularized empirical risk minimization / Strongly convex function / / /  
Reference Info. IEICE Tech. Rep., vol. 122, no. 405, IBISML2022-130, pp. 347-354, March 2023.
Paper # IBISML2022-130 
Date of Issue 2023-02-23 (PRMU, IBISML) 
ISSN 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 PRMU2022-123 IBISML2022-130

Conference Information
Committee PRMU IBISML IPSJ-CVIM  
Conference Date 2023-03-02 - 2023-03-03 
Place (in Japanese) (See Japanese page) 
Place (in English) Future University Hakodate 
Topics (in Japanese) (See Japanese page) 
Topics (in English)  
Paper Information
Registration To IBISML 
Conference Code 2023-03-PRMU-IBISML-CVIM 
Language Japanese 
Title (in Japanese) (See Japanese page) 
Sub Title (in Japanese) (See Japanese page) 
Title (in English) Fast Identification of Possible Model Parameter Update for Low-Rank Update of Training Data 
Sub Title (in English)  
Keyword(1) Low-rank update of data  
Keyword(2) Leave-one-out cross-validation  
Keyword(3) Stepwise feature selection  
Keyword(4) Regularized empirical risk minimization  
Keyword(5) Strongly convex function  
Keyword(6)  
Keyword(7)  
Keyword(8)  
1st Author's Name Hiroyuki Hanada  
1st Author's Affiliation RIKEN (RIKEN)
2nd Author's Name Noriaki Hashimoto  
2nd Author's Affiliation RIKEN (RIKEN)
3rd Author's Name Kouichi Taji  
3rd Author's Affiliation Nagoya University (Nagoya Univ.)
4th Author's Name Ichiro Takeuchi  
4th Author's Affiliation Nagoya University (Nagoya 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 2023-03-03 16:25:00 
Presentation Time 15 minutes 
Registration for IBISML 
Paper # PRMU2022-123, IBISML2022-130 
Volume (vol) vol.122 
Number (no) no.404(PRMU), no.405(IBISML) 
Page pp.347-354 
#Pages
Date of Issue 2023-02-23 (PRMU, IBISML) 


[Return to Top Page]

[Return to IEICE Web Page]


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