Learning kernels for visual domain adaptation

Statistical machine learning has become an important driving force behind many application fields. By large, however, its theoretical underpinning has hinged on the stringent assumption that the learn