| Presentation | 2017-09-15 On MDL Learning of Gaussian Mixture Modlels Kohei Miyamoto, Masanori Kawakita, Jun'ichi Takeuchi, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | The final goal of this work is model sellection for gaussian mixture models(GMM) based on the minimum description length(MDL) principle. In this paper, we discuss some issues about a two part coding for GMM for this purpose. First we consider mixture models which have known distributions as their components. We introduce a two part codes for that mixture models and derive a risk bound for typical sequences and a probabilistic loss bound when the Renyi divergence is used as a loss function. Next we examine their Fisher information. its determinant and Jeffrey's prior which is related to the restriction on parameters. These works will be useful when we consider a two part coding of GMM itself. Further, we study relation between Rissanen's $g$-function and Fisher information. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | MDL / Gaussian mixture |
| Paper # | PRMU2017-47,IBISML2017-19 |
| Date of Issue | 2017-09-08 (PRMU, IBISML) |
| Conference Information | |
| Committee | PRMU / IBISML / IPSJ-CVIM |
|---|---|
| Conference Date | 2017/9/15(2days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | |
| Chair | Shinichi Sato(NII) / Kenji Fukumizu(ISM) |
| Vice Chair | Hironobu Fujiyoshi(Chubu Univ.) / Yoshihisa Ijiri(Omron) / Masashi Sugiyama(Univ. of Tokyo) |
| Secretary | Hironobu Fujiyoshi(AIST) / Yoshihisa Ijiri(NAIST) / Masashi Sugiyama(Kyoto Univ.) / (Univ. of Tokyo) |
| Assistant | Masato Ishii(NEC) / Yusuke Sugano(Osaka Univ.) / Ichiro Takeuchi(Nagoya Inst. of Tech.) / Toshihiro Kamishima(AIST) |
| Paper Information | |
| Registration To | Technical Committee on Pattern Recognition and Media Understanding / Technical Committee on Infomation-Based Induction Sciences and Machine Learning / Special Interest Group on Computer Vision and Image Media |
|---|---|
| Language | JPN |
| Title (in Japanese) | (See Japanese page) |
| Sub Title (in Japanese) | (See Japanese page) |
| Title (in English) | On MDL Learning of Gaussian Mixture Modlels |
| Sub Title (in English) | |
| Keyword(1) | MDL |
| Keyword(2) | Gaussian mixture |
| 1st Author's Name | Kohei Miyamoto |
| 1st Author's Affiliation | Kyushu University(Kyushu Univ.) |
| 2nd Author's Name | Masanori Kawakita |
| 2nd Author's Affiliation | Kyushu University(Kyushu Univ.) |
| 3rd Author's Name | Jun'ichi Takeuchi |
| 3rd Author's Affiliation | Kyushu University(Kyushu Univ.) |
| Date | 2017-09-15 |
| Paper # | PRMU2017-47,IBISML2017-19 |
| Volume (vol) | vol.117 |
| Number (no) | PRMU-210,IBISML-211 |
| Page | pp.pp.59-66(PRMU), pp.59-66(IBISML), |
| #Pages | 8 |
| Date of Issue | 2017-09-08 (PRMU, IBISML) |