Hyperkernel Based Density Estimation

Nikolaos Vasiloglou

Abstract:

We focus on solving the problem of learning an optimal smoothing kernel for the
unsupervised learning problem of kernel density estimation(KDE) by using hyperkernels. The optimal kernel is the one which minimizes the regularized negative
leave-one-out-log likelihood score of the train set. We demonstrate that ”fixed
bandwidth” and ”variable bandwidth” KDE are special cases of our algorithm.