Permutation Complexity Bound on Out-Sample Error

Abstract

We define a data dependent permutation complexity for a hypothesis set , which is similar to a Rademacher complexity or maximum discrepancy. The permutation complexity is based like the maximum discrepancy on (dependent) sampling. We prove a uniform bound on the generalization error, as well as a concentration result which means that the permutation estimate can be efficiently estimated.

Tasks

Reproductions