Presentation | 2010-01-19 An Extension of Matrix Factorization to Mixture Model and an Application to Prediction of Movei Ratings Masayoshi NAKAMURA, Takashi TAKENOUCHI, Kazushi IKEDA, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Recommendation systems suggest items to an user, based on existing history of evaluation for items by the users. In many cases, history information of users is represented as a matrix form, in which most part are missing. Matrix factorization is a well-known method for predicting missing values of the sparse matrix. The cost function of the conventional matrix factorization corresponds to the log-likelihood with a single Gaussian model, whose mean is assumed to be factorizes. However, in real-world datasets, a group of users forms a cluster and each cluster has different characteristic. For the situation the model of single Gaussian is not appropriate. To tackle the situation, in this report, we extend matrix factorization with a mixture model, and investigate our method using a synthetic dataset and a real-world dataset. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Matrix factorization / Mixture model / Missing value prediction / Probabilistic model |
Paper # | NC2009-86 |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 2010/1/11(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 | Neurocomputing (NC) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | An Extension of Matrix Factorization to Mixture Model and an Application to Prediction of Movei Ratings |
Sub Title (in English) | |
Keyword(1) | Matrix factorization |
Keyword(2) | Mixture model |
Keyword(3) | Missing value prediction |
Keyword(4) | Probabilistic model |
1st Author's Name | Masayoshi NAKAMURA |
1st Author's Affiliation | Graduate School of Information Science, Nara Institute of Science and Technology() |
2nd Author's Name | Takashi TAKENOUCHI |
2nd Author's Affiliation | Graduate School of Information Science, Nara Institute of Science and Technology |
3rd Author's Name | Kazushi IKEDA |
3rd Author's Affiliation | Graduate School of Information Science, Nara Institute of Science and Technology |
Date | 2010-01-19 |
Paper # | NC2009-86 |
Volume (vol) | vol.109 |
Number (no) | 363 |
Page | pp.pp.- |
#Pages | 5 |
Date of Issue |