EXTRACTING FEATURE SUBSPACE FOR KERNEL BASED LINEAR PROGRAMMING SUPPORT VECTOR MACHINES
-
- Yajima Yasutoshi
- Tokyo Institute of Technology
-
- Ohi Hiroko
- Hitachi, Ltd.
-
- Mori Masao
- Keio University
この論文をさがす
説明
We propose linear programming formulations of support vector machines (SVM). Unlike standard SVMs which use quadratic programs, our approach explores a fairly small dimensional subspace of a feature space to construct the nonlinear discriminator. This allows us to obtain the discriminator by solving a smaller sized linear program. We demonstrate that an orthonormal basis of the subspace can be implicitly treated by eigenvectors of the Gram matrix defined by the associated kernel function. When the number of given data points is very large, we construct a subspace by random sampling of data points. Numerical experiments indicate that the subspace generated by less than 2% of the entire training data points achieves reasonable performance for a fairly large instance with 60000 data points.
収録刊行物
-
- 日本オペレーションズ・リサーチ学会論文誌
-
日本オペレーションズ・リサーチ学会論文誌 46 (4), 395-408, 2003
公益社団法人 日本オペレーションズ・リサーチ学会
- Tweet
キーワード
詳細情報 詳細情報について
-
- CRID
- 1390282679087502208
-
- NII論文ID
- 110001183580
-
- NII書誌ID
- AA00703935
-
- ISSN
- 21888299
- 04534514
-
- NDL書誌ID
- 6788919
-
- 本文言語コード
- en
-
- データソース種別
-
- JaLC
- NDLサーチ
- Crossref
- CiNii Articles
- KAKEN
- OpenAIRE
-
- 抄録ライセンスフラグ
- 使用不可