Presentation | 2010-11-04 An Application of Generalized Linear Model for Recommender System : Rating Estimation Based on Main-Effect Model Yu FUJIMOTO, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Collaborative filtering based on a rating matrix is broadly used in recommender systems. In a practical situation, the matrix tends to be sparse when the sets of items and objects are huge. And, sparsity of rating matrices easily deteriorate the accuracy of recommendation. In the regression setup, sparse matrices cause the over-fitting problem, and the number of parameters in a model should be well controlled. In this paper, a simple main-effect model with a small number of parameters is introduced and extended in the framework of the generalized linear model. Even with such a simple model, one can express linearity, independence, and weak special types of non-linearity and dependence between users and objects by introducing a one-parameter family link function. This paper experimentally shows a possibility for improvement of estimation results based on a simple model. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | Generalized linear model / collaborative filter / recommender system / linearity, independence |
Paper # | IBISML2010-68 |
Date of Issue |
Conference Information | |
Committee | IBISML |
---|---|
Conference Date | 2010/10/28(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 | Information-Based Induction Sciences and Machine Learning (IBISML) |
---|---|
Language | JPN |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | An Application of Generalized Linear Model for Recommender System : Rating Estimation Based on Main-Effect Model |
Sub Title (in English) | |
Keyword(1) | Generalized linear model |
Keyword(2) | collaborative filter |
Keyword(3) | recommender system |
Keyword(4) | linearity, independence |
1st Author's Name | Yu FUJIMOTO |
1st Author's Affiliation | Aoyama Gakuin University() |
Date | 2010-11-04 |
Paper # | IBISML2010-68 |
Volume (vol) | vol.110 |
Number (no) | 265 |
Page | pp.pp.- |
#Pages | 7 |
Date of Issue |