AI

Adaptation Based on Generalized Discrepancy

Abstract

We present a new algorithm for domain adaptation improving upon a discrepancy mini- mization algorithm, (DM), previously shown to outperform a number of algorithms for this problem. Unlike many previous proposed solutions for domain adaptation, our algorithm does not consist of a fixed reweighting of the losses over the training sample. Instead, the reweighting depends on the hypothesis sought. The algorithm is derived from a less con- servative notion of discrepancy than the DM algorithm. We call this quantity generalized discrepancy. We present a detailed description of our algorithm and show that it can be formulated as a convex optimization problem. We also give a detailed theoretical analysis of its learning guarantees which helps us select its parameters. Finally, we report the results of experiments demonstrating that it improves upon discrepancy minimization in several tasks.