Publica
Hier finden Sie wissenschaftliche Publikationen aus den FraunhoferInstituten. Model selection under covariate shift
 Duch, W.; Kacprzyk, J.; Oja, E.; Zadrozny, S.: Artificial neural networks: formal models and their applications  ICANN 2005 : 15th international conference, Warsaw, Poland, September 1115, 2005. Proceedings. Pt. 2 Berlin: Springer, 2005 (Lecture Notes in Computer Science 3697) ISBN: 3540287558 ISBN: 9783540287551 pp.235240 
 International Conference on Artificial Neural Networks (ICANN) <15, 2005, Warsaw> 

 English 
 Conference Paper 
 Fraunhofer FIRST () 
Abstract
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 assumptionknown as the covariate shiftcauses a heavy bias in standard generalization error estimation schemes such as crossvalidation and thus they result in poor model selection. In this paper, we therefore propose an alternative estimator of the generalization error. Under covariate shift, the proposed generalization error estimator is unbiased if the learning target function is included in the model at hand and it is asymptotically unbiased in general. Experimental results show that model selection with the proposed generalization error estimator is compared favorably to crossvalidation in extrapolation.