Localized Multiple Kernel Machines for Image Recognition
Mehmet Gönen, Ethem Alpaydn · 2009
We review our work on localized multiple kernels (Gonen and Alpaydn, 2008, 2009) that allows kernels to be combined with dierent weights in dierent regions of the input space by using a gating model. We give example uses in image recognition for combining kernels of dierent representations and costs.