Presentation 2022-03-08
Robust computation of optimal transport by β-potential regularization
Shintaro Nakamura, Han Bao, Masashi Sugiyama,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Optimal transport (OT) has become a widely used tool to measure the discrepancy between probability distributionsin the machine learning field. For instance, OT is a popular loss function that quantifies the discrepancy between an empiricaldistribution and a parametric model. Recently, an entropic penalty term and the celebrated Sinkhorn algorithm have beencommonly used to approximate the original OT in a computationally efficient way. However, since the Sinkhorn algorithm runsa projection associated with the Kullback-Leibler divergence, it is often vulnerable to outliers. To overcome this problem, wepropose regularizing OT with the
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Optimal transportRobustnessoutlier detection
Paper # IBISML2021-31
Date of Issue 2022-03-01 (IBISML)

Conference Information
Committee IBISML
Conference Date 2022/3/8(2days)
Place (in Japanese) (See Japanese page)
Place (in English) Online
Topics (in Japanese) (See Japanese page)
Topics (in English) Machine Learning, etc.
Chair Ichiro Takeuchi(Nagoya Inst. of Tech.)
Vice Chair Masashi Sugiyama(Univ. of Tokyo)
Secretary Masashi Sugiyama(Univ. of Tokyo)
Assistant Tomoharu Iwata(NTT) / Atsuyoshi Nakamura(Hokkaido Univ.)

Paper Information
Registration To Technical Committee on Infomation-Based Induction Sciences and Machine Learning
Language ENG-JTITLE
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Robust computation of optimal transport by β-potential regularization
Sub Title (in English)
Keyword(1) Optimal transportRobustnessoutlier detection
1st Author's Name Shintaro Nakamura
1st Author's Affiliation University of Tokyo(Univ. Tokyo)
2nd Author's Name Han Bao
2nd Author's Affiliation University of Tokyo/RIKEN(Univ.Tokyo/RIKEN)
3rd Author's Name Masashi Sugiyama
3rd Author's Affiliation RIKEN/University of Tokyo(RIKEN/Univ. Tokyo)
Date 2022-03-08
Paper # IBISML2021-31
Volume (vol) vol.121
Number (no) IBISML-419
Page pp.pp.8-14(IBISML),
#Pages 7
Date of Issue 2022-03-01 (IBISML)