| Presentation | 2010-12-10 Comparison between Minimum Classification Error Training and a Support Vector Machine in Prototype-based Classifier Design Mamoru ADACHI, Hideyuki WATANABE, Shigeru KATAGIRI, Miho OHSAKI, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | In this paper, we experimentally compared the size of trained classifier class models (prototypes) between Support Vector Machine (SVM) and a newly proposed discriminative training method called Large Geometric Margin Minimum Classification Error (LGM-MCE) training. We first clarify that SVM often requires a large number of prototypes (i.e., support vectors), which result in a scalability problem. Next we show the insufficiency of SVM when the number of support vectors is limited and clearly demonstrate that LGM-MCE is much more useful for realizing a small-size but highly-discriminative classifier than SVM. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Minimum classification error training / Support Vector Machine / Prototype-based classifier / K-means clustering |
| Paper # | PRMU2010-146 |
| Date of Issue |
| Conference Information | |
| Committee | PRMU |
|---|---|
| Conference Date | 2010/12/2(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 | Pattern Recognition and Media Understanding (PRMU) |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | Comparison between Minimum Classification Error Training and a Support Vector Machine in Prototype-based Classifier Design |
| Sub Title (in English) | |
| Keyword(1) | Minimum classification error training |
| Keyword(2) | Support Vector Machine |
| Keyword(3) | Prototype-based classifier |
| Keyword(4) | K-means clustering |
| 1st Author's Name | Mamoru ADACHI |
| 1st Author's Affiliation | Doshisha University() |
| 2nd Author's Name | Hideyuki WATANABE |
| 2nd Author's Affiliation | MASTAR Project, NICT |
| 3rd Author's Name | Shigeru KATAGIRI |
| 3rd Author's Affiliation | Doshisha University |
| 4th Author's Name | Miho OHSAKI |
| 4th Author's Affiliation | Doshisha University |
| Date | 2010-12-10 |
| Paper # | PRMU2010-146 |
| Volume (vol) | vol.110 |
| Number (no) | 330 |
| Page | pp.pp.- |
| #Pages | 6 |
| Date of Issue | |