Conditional Bures Metric for Domain Adaptation

TL;DR

This paper introduces the Conditional Kernel Bures (CKB) metric to measure and align conditional distributions in unsupervised domain adaptation, improving transfer learning by preserving discriminative information.

Contribution

It proposes a novel CKB metric for conditional distribution discrepancy and develops a conditional distribution matching network for better domain adaptation.

Findings

The CKB metric effectively characterizes conditional distribution shifts.

The proposed model outperforms existing UDA methods in experiments.

Theoretical guarantees ensure the convergence of the empirical estimation.

Abstract

As a vital problem in classification-oriented transfer, unsupervised domain adaptation (UDA) has attracted widespread attention in recent years. Previous UDA methods assume the marginal distributions of different domains are shifted while ignoring the discriminant information in the label distributions. This leads to classification performance degeneration in real applications. In this work, we focus on the conditional distribution shift problem which is of great concern to current conditional invariant models. We aim to seek a kernel covariance embedding for conditional distribution which remains yet unexplored. Theoretically, we propose the Conditional Kernel Bures (CKB) metric for characterizing conditional distribution discrepancy, and derive an empirical estimation for the CKB metric without introducing the implicit kernel feature map. It provides an interpretable approach to understand the knowledge transfer mechanism. The established consistency theory of the empirical estimation provides a theoretical guarantee for convergence. A conditional distribution matching network is proposed to learn the conditional invariant and discriminative features for UDA. Extensive experiments and analysis show the superiority of our proposed model.