Asymptotic Bayesian generalization error when training and test distributions are different

Keisuke Yamazaki, Motoaki Kawanabe, Sumio Watanabe, Masashi Sugiyama, Klaus Robert Müller

Research output: Contribution to conferencePaperpeer-review

28 Citations (Scopus)

Abstract

In supervised learning, we commonly assume that training and test data are sampled from the same distribution. However, this assumption can be violated in practice and then standard machine learning techniques perform poorly. This paper focuses on revealing and improving the performance of Bayesian estimation when the training and test distributions are different. We formally analyze the asymptotic Bayesian generalization error and establish its upper bound under a very general setting. Our important finding is that lower order terms - -which can be ignored in the absence of the distribution change - -play an important role under the distribution change. We also propose a novel variant of stochastic complexity which can be used for choosing an appropriate model and hyper-parameters under a particular distribution change.

Original languageEnglish
Pages1079-1086
Number of pages8
DOIs
Publication statusPublished - 2007
Event24th International Conference on Machine Learning, ICML 2007 - Corvalis, OR, United States
Duration: 2007 Jun 202007 Jun 24

Other

Other24th International Conference on Machine Learning, ICML 2007
Country/TerritoryUnited States
CityCorvalis, OR
Period07/6/2007/6/24

ASJC Scopus subject areas

  • Software
  • Human-Computer Interaction
  • Computer Vision and Pattern Recognition
  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Asymptotic Bayesian generalization error when training and test distributions are different'. Together they form a unique fingerprint.

Cite this