A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning

TLDR

Sequential prediction tasks such as imitation learning violate i.i.d. assumptions, leading to poor theoretical and practical performance; recent methods offer stronger guarantees but are limited by non‑stationary or stochastic policies and many iterations. The authors aim to develop an iterative algorithm that trains a stationary deterministic policy as a no‑regret online learner. The algorithm is formulated as a no‑regret online learning procedure that iteratively updates the policy.

Abstract

Sequential prediction problems such as imitation learning, where future observations depend on previous predictions (actions), violate the common i.i.d. assumptions made in statistical learning. This leads to poor performance in theory and often in practice. Some recent approaches provide stronger guarantees in this setting, but remain somewhat unsatisfactory as they train either non-stationary or stochastic policies and require a large number of iterations. In this paper, we propose a new iterative algorithm, which trains a stationary deterministic policy, that can be seen as a no regret algorithm in an online learning setting. We show that any such no regret algorithm, combined with additional reduction assumptions, must find a policy with good performance under the distribution of observations it induces in such sequential settings. We demonstrate that this new approach outperforms previous approaches on two challenging imitation learning problems and a benchmark sequence labeling problem.