Regret Bounds for Reinforcement Learning with Policy Advice
This provides a method for efficiently leveraging prior policies in large reinforcement learning domains, though it is incremental as it builds on existing regret minimization frameworks.
The paper tackles the problem of reinforcement learning when a set of input policies is available, by proposing an algorithm (RLPA) that learns to use the best policy in the set, achieving a sub-linear regret of ˜O(√T) relative to that policy, with regret and computational complexity independent of state and action space sizes.
In some reinforcement learning problems an agent may be provided with a set of input policies, perhaps learned from prior experience or provided by advisors. We present a reinforcement learning with policy advice (RLPA) algorithm which leverages this input set and learns to use the best policy in the set for the reinforcement learning task at hand. We prove that RLPA has a sub-linear regret of \tilde O(\sqrt{T}) relative to the best input policy, and that both this regret and its computational complexity are independent of the size of the state and action space. Our empirical simulations support our theoretical analysis. This suggests RLPA may offer significant advantages in large domains where some prior good policies are provided.