Combining predictions from linear models when training and test inputs differ.

Thijs van Ommen

Browse the full UAI paper archive.