Provably Breaking the Quadratic Error Compounding Barrier in Imitation Learning, Optimally
share
We study the statistical limits of Imitation Learning (IL) in episodic Markov Decision Processes (MDPs) with a state space 𝒮. We focus on the known-transition setting where the learner is provided a dataset of N length-H trajectories from a deterministic expert policy and knows the MDP transition. We establish an upper bound O(|𝒮|H^3/2/N) for the suboptimality using the Mimic-MD algorithm in Rajaraman et al (2020) which we prove to be computationally efficient. In contrast, we show the minimax suboptimality grows as Ω( H^3/2/N) when |𝒮|≥ 3 while the unknown-transition setting suffers from a larger sharp rate Θ(|𝒮|H^2/N) (Rajaraman et al (2020)). The lower bound is established by proving a two-way reduction between IL and the value estimation problem of the unknown expert policy under any given reward function, as well as building connections with linear functional estimation with subsampled observations. We further show that under the additional assumption that the expert is optimal for the true reward function, there exists an efficient algorithm, which we term as Mimic-Mixture, that provably achieves suboptimality O(1/N) for arbitrary 3-state MDPs with rewards only at the terminal layer. In contrast, no algorithm can achieve suboptimality O(√(H)/N) with high probability if the expert is not constrained to be optimal. Our work formally establishes the benefit of the expert optimal assumption in the known transition setting, while Rajaraman et al (2020) showed it does not help when transitions are unknown.
READ FULL TEXT