Feature-Level Domain Adaptation

TL;DR

This paper introduces feature-level domain adaptation (FLDA), a method that models the transfer between source and target domains at the feature level, enabling efficient adaptation of classifiers to distribution shifts.

Contribution

The paper proposes FLDA, a novel approach that models domain transfer at the feature level and provides an analytical framework for efficient classifier adaptation.

Findings

FLDA performs comparably to state-of-the-art methods.

It effectively models transfer using dropout distributions.

Efficient analytical computation of expected loss is achieved.

Abstract

Domain adaptation is the supervised learning setting in which the training and test data are sampled from different distributions: training data is sampled from a source domain, whilst test data is sampled from a target domain. This paper proposes and studies an approach, called feature-level domain adaptation (FLDA), that models the dependence between the two domains by means of a feature-level transfer model that is trained to describe the transfer from source to target domain. Subsequently, we train a domain-adapted classifier by minimizing the expected loss under the resulting transfer model. For linear classifiers and a large family of loss functions and transfer models, this expected loss can be computed or approximated analytically, and minimized efficiently. Our empirical evaluation of FLDA focuses on problems comprising binary and count data in which the transfer can be naturally modeled via a dropout distribution, which allows the classifier to adapt to differences in the marginal probability of features in the source and the target domain. Our experiments on several real-world problems show that FLDA performs on par with state-of-the-art domain-adaptation techniques.