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 language | English |
---|---|
Pages | 1079-1086 |
Number of pages | 8 |
DOIs | |
Publication status | Published - 2007 |
Event | 24th International Conference on Machine Learning, ICML 2007 - Corvalis, OR, United States Duration: 2007 Jun 20 → 2007 Jun 24 |
Other
Other | 24th International Conference on Machine Learning, ICML 2007 |
---|---|
Country/Territory | United States |
City | Corvalis, OR |
Period | 07/6/20 → 07/6/24 |
ASJC Scopus subject areas
- Software
- Human-Computer Interaction
- Computer Vision and Pattern Recognition
- Computer Networks and Communications