The Application of Deep Kernel Machines to Various Types of Data
Xiaohui Wang · 2011
Typically, kernel machines are linear classifiers in the implicit feature space. We argue that linear classification in the kernel’s implicit feature space may sometimes be inadequate, especially in situations where the choices of kernel functions are limited. When this is the case, one naturally considers nonlinear classifiers in the feature space. We show that repeating this process produces something we call deep kernel machines. We apply this new algorithm to a various of data with different types of kernels. Results show that these deep kernel machines can make a tangible difference in classification performance.