NAS-OoD: Neural Architecture Search for Out-of-Distribution Generalization

TL;DR

NAS-OoD introduces a novel neural architecture search method that optimizes architectures for out-of-distribution generalization by jointly learning a data generator and architecture parameters through a minimax training process, leading to improved robustness.

Contribution

The paper proposes NAS-OoD, a new approach that jointly optimizes neural architectures and a data generator for OoD generalization, addressing the limitations of traditional NAS methods.

Findings

Achieves superior OoD generalization performance on benchmarks.

Reduces error rate by over 70% on a real industry dataset.

Uses fewer parameters while maintaining robustness.

Abstract

Recent advances on Out-of-Distribution (OoD) generalization reveal the robustness of deep learning models against distribution shifts. However, existing works focus on OoD algorithms, such as invariant risk minimization, domain generalization, or stable learning, without considering the influence of deep model architectures on OoD generalization, which may lead to sub-optimal performance. Neural Architecture Search (NAS) methods search for architecture based on its performance on the training data, which may result in poor generalization for OoD tasks. In this work, we propose robust Neural Architecture Search for OoD generalization (NAS-OoD), which optimizes the architecture with respect to its performance on generated OoD data by gradient descent. Specifically, a data generator is learned to synthesize OoD data by maximizing losses computed by different neural architectures, while the goal for architecture search is to find the optimal architecture parameters that minimize the synthetic OoD data losses. The data generator and the neural architecture are jointly optimized in an end-to-end manner, and the minimax training process effectively discovers robust architectures that generalize well for different distribution shifts. Extensive experimental results show that NAS-OoD achieves superior performance on various OoD generalization benchmarks with deep models having a much fewer number of parameters. In addition, on a real industry dataset, the proposed NAS-OoD method reduces the error rate by more than 70% compared with the state-of-the-art method, demonstrating the proposed method's practicality for real applications.