Learning Markov State Abstractions For Deep Reinforcement Learning

Abstract

A fundamental assumption of reinforcement learning in Markov decision processes (MDPs) is that the relevant decision process is, in fact, Markov. However, when MDPs have rich observations, agents typically learn by way of an abstract state representation, and such representations are not guaranteed to preserve the Markov property. We introduce a novel set of conditions and prove that they are sufficient for learning a Markov abstract state representation. We then describe a practical training procedure that combines inverse model estimation and temporal contrastive learning to learn an abstraction that approximately satisfies these conditions. Our novel training objective is compatible with both online and offline training: it does not require a reward signal, but agents can capitalize on reward information when available. We empirically evaluate our approach on a visual gridworld domain and a set of continuous control benchmarks. Our approach learns representations that capture the un

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyallen2021learning

Related papers

• Markov Abstractions For PAC Reinforcement Learning In Non-markov Decision Processes (2022)0.00
• An Analysis Of Model-based Reinforcement Learning From Abstracted Observations (2022)0.00
• On Learning History Based Policies For Controlling Markov Decision Processes (2022)0.00
• Contrastive Abstraction For Reinforcement Learning (2024)0.00
• Deepmdp: Learning Continuous Latent Space Models For Representation Learning (2019)0.00
• Learning Robust State Abstractions For Hidden-parameter Block Mdps (2020)0.00
• Learning Non-markovian Reward Models In Mdps (2020)0.00
• Bridging State And History Representations: Understanding Self-predictive RL (2024)0.00