Presentation 2005/3/23
Incremental learning and forgetting using Sensitivity in SVM
Takeshi ASADA, Hirotaka NAKAYAMA, Tetsuzo TANINO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Recently, support vector machine (SVM) is one of the hot topics in the field of pattern recognition. There are some studies to apply SVM to incremental and decremental learning. Chunking method is one of the incremental learning model to improve computational time. Conventional chunking method sometimes obtains low generalization. In order to overcome this problem, in this paper, chunking method model considering sensitivity is proposed. By the way, forgetting requires to reduce the influence of the data. Active forgetting, in particular, requires to reduce the influence of obstacle data. In this paper, to select obstacle data, active forgetting model considering sensitivity is proposed.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Support vector machines / Sensitivity / Incremental learning / Chunking method / Convex hull / Forgetting
Paper # NC2004-222
Date of Issue

Conference Information
Committee NC
Conference Date 2005/3/23(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Incremental learning and forgetting using Sensitivity in SVM
Sub Title (in English)
Keyword(1) Support vector machines
Keyword(2) Sensitivity
Keyword(3) Incremental learning
Keyword(4) Chunking method
Keyword(5) Convex hull
Keyword(6) Forgetting
1st Author's Name Takeshi ASADA
1st Author's Affiliation Graduate School of Engineering, Osaka University()
2nd Author's Name Hirotaka NAKAYAMA
2nd Author's Affiliation Department of Information Science and Systems Engineering, Konan University
3rd Author's Name Tetsuzo TANINO
3rd Author's Affiliation Graduate School of Engineering, Osaka University
Date 2005/3/23
Paper # NC2004-222
Volume (vol) vol.104
Number (no) 760
Page pp.pp.-
#Pages 6
Date of Issue