Kernelized Heterogeneous Risk Minimization

Abstract

The ability to generalize under distributional shifts is essential to reliable machine learning, while models optimized with empirical risk minimization usually fail on non-$i.i.d$ testing data. Recently, invariant learning methods for out-of-distribution (OOD) generalization propose to find causally invariant relationships with multi-environments. However, modern datasets are frequently multi-sourced without explicit source labels, rendering many invariant learning methods inapplicable. In this paper, we propose Kernelized Heterogeneous Risk Minimization (KerHRM) algorithm, which achieves both the latent heterogeneity exploration and invariant learning in kernel space, and then gives feedback to the original neural network by appointing invariant gradient direction. We theoretically justify our algorithm and empirically validate the effectiveness of our algorithm with extensive experiments.

arXiv comment35th Conference on Neural Information Processing Systems (NeurIPS 2021), Sydney, Australia. arXiv admin note: text overlap with arXiv:2105.03818

References in corpus (8)

• Neural Tangent Kernel: Convergence and Generalization in Neural Networks
• Domain Adaptation for Statistical Classifiers
• Distributionally Robust Neural Networks for Group Shifts: On the Importance of Regularization for Worst-Case Generalization
• Does Distributionally Robust Supervised Learning Give Robust Classifiers?
• Generalization Guarantees for Neural Networks via Harnessing the Low-rank Structure of the Jacobian
• Incorporating Unlabeled Data into Distributionally Robust Learning
• Environment Inference for Invariant Learning
• Heterogeneous Risk Minimization

Cited by in corpus (2)

• Towards out of distribution generalization for problems in mechanics
• Pareto Invariant Representation Learning for Multimedia Recommendation