Efficient Reinforcement Learning in Block MDPs: A Model-free Representation Learning Approach
This addresses the problem of sample-efficient RL in complex environments with rich observations for AI and robotics applications, representing a novel method for a known bottleneck.
The paper tackles efficient reinforcement learning in Block MDPs with rich observations by proposing BRIEE, which interleaves latent state discovery, exploration, and exploitation, achieving provable near-optimal policy learning with polynomial sample complexity independent of observation space size and showing empirical improvements over state-of-the-art methods on combination lock problems.
We present BRIEE (Block-structured Representation learning with Interleaved Explore Exploit), an algorithm for efficient reinforcement learning in Markov Decision Processes with block-structured dynamics (i.e., Block MDPs), where rich observations are generated from a set of unknown latent states. BRIEE interleaves latent states discovery, exploration, and exploitation together, and can provably learn a near-optimal policy with sample complexity scaling polynomially in the number of latent states, actions, and the time horizon, with no dependence on the size of the potentially infinite observation space. Empirically, we show that BRIEE is more sample efficient than the state-of-art Block MDP algorithm HOMER and other empirical RL baselines on challenging rich-observation combination lock problems that require deep exploration.