Sparse support vector machine for big problem which sequentially-add categories

2014 Fiscal Year Final Research Report

• Project/Area Number: 25871033
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Intelligent informatics; Perceptual information processing
• Research Institution: National Institute of Technology, Toyama College
• Principal Investigator: KITAMURA Takuya 富山高等専門学校, 電気制御システム工学科, 助教 (40611918)
• Project Period (FY): 2013-04-01 – 2015-03-31
• Keywords: パターン認識 / 機械学習 / サポートベクトルマシン / スパース / オンライン学習

Outline of Final Research Achievements

In this research, I have developed sparse support vector machine for big problem which sequentially-add categories. For example, I apply this system to face identification which sequentially-add the categories (registrants). Then, this problem may be too big. However, this system uses only a new added category in training when this category is added. Using face identification problem and multi-category benchmark datasets, I evaluate the effectiveness of this system.

Free Research Field

機械学習