Presentation | 2005/3/23 Generalization Error Estimation When Training and Test Input Points Follow Different Probability Distributions Masashi SUGIYAMA, Klaus Robert MULLER, |
---|---|
PDF Download Page | PDF download Page Link |
Abstract(in Japanese) | (See Japanese page) |
Abstract(in English) | A common assumption in supervised learning is that the training and test input points follow the same probability distribution. However, this assumption is not fulfilled, e.g., in interpolation, extrapolation, or active learning scenarios. The violation of this assumption-known as the covariate shift-causes a heavy bias in standard generalization error estimation schemes such as cross-validation, and thus they result in poor model selection. In this paper, we therefore propose an alternative estimator of the generalization error. Under the covariate shift, the proposed generalization error estimator is exactly unbiased with finite samples if the learning target function is in the model at hand, and it is asymptotically unbiased in general. We experimentally show that model selection with the proposed generalization error estimator is compared favorably to cross-validation in extrapolation. |
Keyword(in Japanese) | (See Japanese page) |
Keyword(in English) | linear regression / generalization error / model selection / covariate shift / sample selection bias / interpolation / extrapolation / active learning |
Paper # | NC2004-215 |
Date of Issue |
Conference Information | |
Committee | NC |
---|---|
Conference Date | 2005/3/23(1days) |
Place (in Japanese) | (See Japanese page) |
Place (in English) | |
Topics (in Japanese) | (See Japanese page) |
Topics (in English) | |
Chair | |
Vice Chair | |
Secretary | |
Assistant |
Paper Information | |
Registration To | Neurocomputing (NC) |
---|---|
Language | ENG |
Title (in Japanese) | (See Japanese page) |
Sub Title (in Japanese) | (See Japanese page) |
Title (in English) | Generalization Error Estimation When Training and Test Input Points Follow Different Probability Distributions |
Sub Title (in English) | |
Keyword(1) | linear regression |
Keyword(2) | generalization error |
Keyword(3) | model selection |
Keyword(4) | covariate shift |
Keyword(5) | sample selection bias |
Keyword(6) | interpolation |
Keyword(7) | extrapolation |
Keyword(8) | active learning |
1st Author's Name | Masashi SUGIYAMA |
1st Author's Affiliation | Department of Computer Science, Tokyo Institute of Technology() |
2nd Author's Name | Klaus Robert MULLER |
2nd Author's Affiliation | Fraunhofer FIRST.IDA:Department of Computer Science, University of Potsdam |
Date | 2005/3/23 |
Paper # | NC2004-215 |
Volume (vol) | vol.104 |
Number (no) | 760 |
Page | pp.pp.- |
#Pages | 6 |
Date of Issue |