Presentation | 2012-11-07 Feature Selection for LDA using Bayesian Network Ryosuke OHATA, Maomi UENO, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | Recently there has been great interest in topic model analyzing discrete data accompanied by arbitrary features, such as authors. However, previous work in such topic modeling does not take account of feature selection. This paper presents a feature selection method for such topic model. The proposed method selects the feature set involving data generation by clarifying causality between the latent topic and features using bayesian network-based approach. The Advantages of the proposed method are as follows: First, computational time is short because the proposed method can select the feature set without learning latent topic of any combination of features. Second, feature selection accuracy is high because the proposed method has consistency. We demonstrate the effectiveness of the proposed method by simulation. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | topic model / feature selection / bayesian network |
Paper # | IBISML2012-53 |
Date of Issue |
Conference Information | |
Committee | IBISML |
---|---|
Conference Date | 2012/10/31(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) | Feature Selection for LDA using Bayesian Network |
Sub Title (in English) | |
Keyword(1) | topic model |
Keyword(2) | feature selection |
Keyword(3) | bayesian network |
1st Author's Name | Ryosuke OHATA |
1st Author's Affiliation | Graduate School of Information Systems, The University of Electro-Communications() |
2nd Author's Name | Maomi UENO |
2nd Author's Affiliation | Graduate School of Information Systems, The University of Electro-Communications |
Date | 2012-11-07 |
Paper # | IBISML2012-53 |
Volume (vol) | vol.112 |
Number (no) | 279 |
Page | pp.pp.- |
#Pages | 8 |
Date of Issue |