Adversarial Online Multi-Task Reinforcement Learning
We consider the adversarial online multi-task reinforcement learning setting, where in each of K episodes the learner is given an unknown task taken from a finite set of M unknown finite-horizon MDP models. The learner's objective is to minimize its regret with respect to the optimal policy for each task. We assume the MDPs in ℳ are well-separated under a notion of λ-separability, and show that this notion generalizes many task-separability notions from previous works. We prove a minimax lower bound of Ω(K√(DSAH)) on the regret of any learning algorithm and an instance-specific lower bound of Ω(K/λ^2) in sample complexity for a class of uniformly-good cluster-then-learn algorithms. We use a novel construction called 2-JAO MDP for proving the instance-specific lower bound. The lower bounds are complemented with a polynomial time algorithm that obtains Õ(K/λ^2) sample complexity guarantee for the clustering phase and Õ(√(MK)) regret guarantee for the learning phase, indicating that the dependency on K and 1/λ^2 is tight.
READ FULL TEXT