| Presentation | 2018-11-05 [Poster Presentation] Inference for Logitistic Regression Mixture Model with Local Variational Approximation and Study for Variational Free Energy Fumito Nakamura, Ryosuke Konishi, Yasushi Kiyoki, |
|---|---|
| PDF Download Page |  PDF download Page Link |
| Abstract(in Japanese) | (See Japanese page) |
| Abstract(in English) | A logistic regression mixture model (LRMM) is a mixed model of the Logistic regression model, and it is widely used in the field of psychology, sociology and marketing due to its high performance. In a conventional method, the Expectation Maximization (EM) algorithm has been often used to estimate the model. However, the EM algorithm searches the local maximum likelihood estimator, and it is known that the maximum likelihood estimator gives worse performance than the Bayesian approach. In this paper, we propose an algorithm to estimate the LRMM by a Local Variational Approximation (LVA), which is one of the Bayesian approach. Numerical experiments show that the LVA achieves the higher performance than the EM algorithm. Furthermore, we discuss the asymptotic behavior of a variational free energy, which is one of an evaluation index for the LVA. |
| Keyword(in Japanese) | (See Japanese page) |
| Keyword(in English) | Logistic Regression Mixture Model / Local Variational Approximation / Variational Free Energy / Bayesian inference / EM algorithm |
| Paper # | IBISML2018-48 |
| Date of Issue | 2018-10-29 (IBISML) |
| Conference Information | |
| Committee | IBISML |
|---|---|
| Conference Date | 2018/11/5(3days) |
| Place (in Japanese) | (See Japanese page) |
| Place (in English) | Hokkaido Citizens Activites Center (Kaderu 2.7) |
| Topics (in Japanese) | (See Japanese page) |
| Topics (in English) | Information-Based Induction Science Workshop (IBIS2018) |
| Chair | Hisashi Kashima(Kyoto Univ.) |
| Vice Chair | Masashi Sugiyama(Univ. of Tokyo) / Koji Tsuda(Univ. of Tokyo) |
| Secretary | Masashi Sugiyama(Nagoya Inst. of Tech.) / Koji Tsuda(AIST) |
| Assistant | Tomoharu Iwata(NTT) / Shigeyuki Oba(Kyoto Univ.) |
| 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) | [Poster Presentation] Inference for Logitistic Regression Mixture Model with Local Variational Approximation and Study for Variational Free Energy |
| Sub Title (in English) | |
| Keyword(1) | Logistic Regression Mixture Model |
| Keyword(2) | Local Variational Approximation |
| Keyword(3) | Variational Free Energy |
| Keyword(4) | Bayesian inference |
| Keyword(5) | EM algorithm |
| 1st Author's Name | Fumito Nakamura |
| 1st Author's Affiliation | Generic Solution Corporation(Generic Solution) |
| 2nd Author's Name | Ryosuke Konishi |
| 2nd Author's Affiliation | Generic Solution Corporation(Generic Solution) |
| 3rd Author's Name | Yasushi Kiyoki |
| 3rd Author's Affiliation | Keio University(Keio) |
| Date | 2018-11-05 |
| Paper # | IBISML2018-48 |
| Volume (vol) | vol.118 |
| Number (no) | IBISML-284 |
| Page | pp.pp.29-36(IBISML), |
| #Pages | 8 |
| Date of Issue | 2018-10-29 (IBISML) |