| Presentation | 2016-11-17 An Exhaustive Search with Support Vector Machine (ES-SVM) for sparse variable selection Daiki Kawabata, Hiroko Ichikawa, Yasuhiko Igarashi, Kenji Nagata, Satoshi Eifuku, Ryoi Tamura, Masato Okada, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | Nagata et al.(2015) has proposed Exhaustive Search with Support Vector Machine(ES-SVM) which calculates a cross validation error(CVE) by Support Vector Machine for all combinations of variables in order to make the histogram of CVE.In this study, we propose a novel method for sparse variable selection called ESn-SVM for classification.The method assumes sparsity to the true solution and increases the variables which useful for classification from one by one.In order to efficiently perform ESn-SVM, we propose an systematic approach by the replica exchange Monte Carlo (REMC) method and the multiple histogram method. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | variable selection / liner classification / exhaustive search / support vector machine (SVM) / density of states / replica exchange Monte Carlo (REMC) |
| Paper # | IBISML2016-96 |
| Date of Issue | 2016-11-09 (IBISML) |
| Conference Information | |
| Committee | IBISML |
|---|---|
| Conference Date | 2016/11/16(3days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Kyoto Univ. |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Information-Based Induction Science Workshop (IBIS2016) |
| Chair | Kenji Fukumizu(ISM) |
| Vice Chair | Masashi Sugiyama(Univ. of Tokyo) / Hisashi Kashima(Kyoto Univ.) |
| Secretary | Masashi Sugiyama(Univ. of Tokyo) / Hisashi Kashima(Nagoya Inst. of Tech.) |
| Assistant | Toshihiro Kamishima(AIST) / Tomoharu Iwata(NTT) |
| Paper Information | |
| Registration To | Technical Committee on Infomation-Based Induction Sciences and Machine Learning |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | An Exhaustive Search with Support Vector Machine (ES-SVM) for sparse variable selection |
| Sub Title (in English) | |
| Keyword(1) | variable selection |
| Keyword(2) | liner classification |
| Keyword(3) | exhaustive search |
| Keyword(4) | support vector machine (SVM) |
| Keyword(5) | density of states |
| Keyword(6) | replica exchange Monte Carlo (REMC) |
| 1st Author's Name | Daiki Kawabata |
| 1st Author's Affiliation | Univercity of Tokyo(UTokyo) |
| 2nd Author's Name | Hiroko Ichikawa |
| 2nd Author's Affiliation | Tokyo University of Science(TUS) |
| 3rd Author's Name | Yasuhiko Igarashi |
| 3rd Author's Affiliation | Univercity of Tokyo(UTokyo) |
| 4th Author's Name | Kenji Nagata |
| 4th Author's Affiliation | National Institute of Advanced Industrial Science and Technology/Japan Science and Technology Agency /Univercity of Tokyo(AIST/JST/UTokyo) |
| 5th Author's Name | Satoshi Eifuku |
| 5th Author's Affiliation | Univercity of Toyama(Toyama Univ.) |
| 6th Author's Name | Ryoi Tamura |
| 6th Author's Affiliation | Univercity of Toyama(Toyama Univ.) |
| 7th Author's Name | Masato Okada |
| 7th Author's Affiliation | Univercity of Tokyo(UTokyo) |
| Date | 2016-11-17 |
| Paper # | IBISML2016-96 |
| Volume (vol) | vol.116 |
| Number (no) | IBISML-300 |
| Page | pp.pp.361-368(IBISML), |
| #Pages | 8 |
| Date of Issue | 2016-11-09 (IBISML) |