A Conditional Expectation Approach to Model Selection and Active Learning under Covariate Shift
Sugiyama Masashi, Neil Rubens, Müller Klaus-Robert · The MIT Press eBooks · 2008
This chapter, which addresses the problems of model selection and active learning in the conditional expectation framework, is organized as follows. Section 7.2 formulates a linear regression problem with covariate shift. Sections 7.3 and 7.4 introduce a model selection criterion and an active learning criterion, respectively, in the conditional expectation framework and show that they are more advantageous than the full expectation methods in the context of approximate linear regression. Section 7.5 discusses how model selection and active learning can be combined. Concluding remarks and future prospects are given in Section 7.6.