Presentation 2018-11-05
[Poster Presentation] Variational Approximation Accuracy in Non-negative Matrix Factorization
Naoki Hayashi,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) The asymptotic behavior of the variational free energy of the non-negative matrix factorization (NMF) has been elucidated, however, it has not been clarified how well the variational approximation accuracy is with respect to the Bayesian inference. In this study, we resolve the behavior of free energy of Bayesian NMF assuming the same likelihood model and prior distribution as variational Bayesian NMF by using algebraic geometrical method, and their behaviors are theoretically compared.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) non-negative matrix factorization (NMF) / singular model / learning coefficient / real log canonical threshold / variational free energy / free energy / variational Bayesian method / Bayesian inference
Paper # IBISML2018-51
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] Variational Approximation Accuracy in Non-negative Matrix Factorization
Sub Title (in English)
Keyword(1) non-negative matrix factorization (NMF)
Keyword(2) singular model
Keyword(3) learning coefficient
Keyword(4) real log canonical threshold
Keyword(5) variational free energy
Keyword(6) free energy
Keyword(7) variational Bayesian method
Keyword(8) Bayesian inference
1st Author's Name Naoki Hayashi
1st Author's Affiliation NTTDATA Mathematical Systems Inc.(MSI)
Date 2018-11-05
Paper # IBISML2018-51
Volume (vol) vol.118
Number (no) IBISML-284
Page pp.pp.53-60(IBISML),
#Pages 8
Date of Issue 2018-10-29 (IBISML)