Presentation 2018-03-06
The improving method of Singular Bayesian information criterion by analyzing learning coefficients
Sayaka Suzuki, Souta Shina, Miki Aoyagi,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Real data associated with genetic analysis, data mining, image or speech recognition, artificial intelligence, the control of a robot, time series prediction, and so on, are mainly analyzed by the layered neural network, the Boltzmann machine, the reduced rank regression and the normal mixture model, etc. Such hierarchical learning models are non-regular statistical models, which cannot be analyzed using the classic theories of regular statistical models. In this paper, we propose a new methodby improving a singular Bayesian information criterion and by considering characters of learning coefficients. We apply the method to normal mixture modelsand show its effectiveness.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) A singular Bayesian information criterion / a learning coefficient / model selection method / a singular model
Paper # IBISML2017-100
Date of Issue 2018-02-26 (IBISML)

Conference Information
Committee IBISML
Conference Date 2018/3/5(2days)
Place (in Japanese) (See Japanese page)
Place (in English) Nishijin Plaza, Kyushu University
Topics (in Japanese) (See Japanese page)
Topics (in English) Statisitical Mathematics, Machine Learning, Data Mining, etc.
Chair Kenji Fukumizu(ISM)
Vice Chair Masashi Sugiyama(Univ. of Tokyo) / Hisashi Kashima(Kyoto Univ.)
Secretary Masashi Sugiyama(Nagoya Inst. of Tech.) / Hisashi Kashima(Univ. of Tokyo)
Assistant Tomoharu Iwata(NTT) / Toshihiro Kamishima(AIST)

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) The improving method of Singular Bayesian information criterion by analyzing learning coefficients
Sub Title (in English)
Keyword(1) A singular Bayesian information criterion
Keyword(2) a learning coefficient
Keyword(3) model selection method
Keyword(4) a singular model
1st Author's Name Sayaka Suzuki
1st Author's Affiliation Nihon University(Nihon Univ.)
2nd Author's Name Souta Shina
2nd Author's Affiliation Nihon University(Nihon Univ.)
3rd Author's Name Miki Aoyagi
3rd Author's Affiliation Nihon University(Nihon Univ.)
Date 2018-03-06
Paper # IBISML2017-100
Volume (vol) vol.117
Number (no) IBISML-475
Page pp.pp.71-76(IBISML),
#Pages 6
Date of Issue 2018-02-26 (IBISML)