Robust Imitation Learning from Corrupted Demonstrations
This addresses the problem of robust policy learning for AI agents when demonstrations are corrupted, which is incremental as it builds on existing imitation learning methods by adding robustness to outliers.
The paper tackles offline Imitation Learning from corrupted demonstrations, where a constant fraction of data may be noise or outliers, by proposing a robust algorithm using a Median-of-Means objective that guarantees accurate policy estimation. The method achieves competitive results on continuous-control benchmarks, with theoretical analysis showing it maintains similar error scaling and sample complexity as classical Behavior Cloning in expert settings.
We consider offline Imitation Learning from corrupted demonstrations where a constant fraction of data can be noise or even arbitrary outliers. Classical approaches such as Behavior Cloning assumes that demonstrations are collected by an presumably optimal expert, hence may fail drastically when learning from corrupted demonstrations. We propose a novel robust algorithm by minimizing a Median-of-Means (MOM) objective which guarantees the accurate estimation of policy, even in the presence of constant fraction of outliers. Our theoretical analysis shows that our robust method in the corrupted setting enjoys nearly the same error scaling and sample complexity guarantees as the classical Behavior Cloning in the expert demonstration setting. Our experiments on continuous-control benchmarks validate that our method exhibits the predicted robustness and effectiveness, and achieves competitive results compared to existing imitation learning methods.