Domain Adaptation: Overfitting and Small Sample Statistics

TL;DR

This paper addresses the challenge of domain adaptation with limited domain samples, proposing a T-statistic based feature selection method that improves robustness against overfitting and generalizes better to new domains.

Contribution

It introduces a theoretical framework and a practical greedy feature selection algorithm using T-statistics for domain adaptation with small sample sizes.

Findings

T-statistic based selection outperforms classical methods in avoiding overfitting

The approach allows selecting more features than the number of domains

Experimental results validate the robustness of the proposed method

Abstract

We study the prevalent problem when a test distribution differs from the training distribution. We consider a setting where our training set consists of a small number of sample domains, but where we have many samples in each domain. Our goal is to generalize to a new domain. For example, we may want to learn a similarity function using only certain classes of objects, but we desire that this similarity function be applicable to object classes not present in our training sample (e.g. we might seek to learn that "dogs are similar to dogs" even though images of dogs were absent from our training set). Our theoretical analysis shows that we can select many more features than domains while avoiding overfitting by utilizing data-dependent variance properties. We present a greedy feature selection algorithm based on using T-statistics. Our experiments validate this theory showing that our T-statistic based greedy feature selection is more robust at avoiding overfitting than the classical greedy procedure.