Non-stationary Domain Generalization: Theory and Algorithm

TL;DR

This paper addresses the challenge of domain generalization in non-stationary environments, providing theoretical insights and proposing an adaptive invariant representation learning algorithm to improve model robustness across evolving domains.

Contribution

It introduces the first theoretical analysis of non-stationary domain generalization and develops a novel adaptive invariant learning algorithm for better generalization.

Findings

Theoretical upper bounds for model error in non-stationary settings.

The proposed algorithm outperforms existing methods on synthetic and real datasets.

Validation shows improved generalization to unseen evolving domains.

Abstract

Although recent advances in machine learning have shown its success to learn from independent and identically distributed (IID) data, it is vulnerable to out-of-distribution (OOD) data in an open world. Domain generalization (DG) deals with such an issue and it aims to learn a model from multiple source domains that can be generalized to unseen target domains. Existing studies on DG have largely focused on stationary settings with homogeneous source domains. However, in many applications, domains may evolve along a specific direction (e.g., time, space). Without accounting for such non-stationary patterns, models trained with existing methods may fail to generalize on OOD data. In this paper, we study domain generalization in non-stationary environment. We first examine the impact of environmental non-stationarity on model performance and establish the theoretical upper bounds for the model error at target domains. Then, we propose a novel algorithm based on adaptive invariant representation learning, which leverages the non-stationary pattern to train a model that attains good performance on target domains. Experiments on both synthetic and real data validate the proposed algorithm.