Rademacher chaos complexities for learning the kernel problem
Massachusetts Institute of Technology Press (MIT Press)
We develop a novel generalization bound for learning the kernel problem. First, we show that the generalization analysis of the kernel learning problem reduces to investigation of the suprema of the Rademacher chaos process of order 2 over candidate kernels, which we refer to as Rademacher chaos complexity. Next, we show how to estimate the empirical Rademacher chaos complexity by well-established metric entropy integrals and pseudo-dimension of the set of candidate kernels. Our new methodology mainly depends on the principal theory of U-processes and entropy integrals. Finally, we establish satisfactory excess generalization bounds and misclassification error rates for learning gaussian kernels and general radial basis kernels.
Copyright © 2010 The MIT Press
Copyright © 2010 Massachusetts Institute of Technology
Vol. 22 (11), pp. 2858 - 2886
Place of publication