Approximate Information State Based Convergence Analysis Of Recurrent Q-learning

Abstract

In spite of the large literature on reinforcement learning (RL) algorithms for partially observable Markov decision processes (POMDPs), a complete theoretical understanding is still lacking. In a partially observable setting, the history of data available to the agent increases over time so most practical algorithms either truncate the history to a finite window or compress it using a recurrent neural network leading to an agent state that is non-Markovian. In this paper, it is shown that in spite of the lack of the Markov property, recurrent Q-learning (RQL) converges in the tabular setting. Moreover, it is shown that the quality of the converged limit depends on the quality of the representation which is quantified in terms of what is known as an approximate information state (AIS). Based on this characterization of the approximation error, a variant of RQL with AIS losses is presented. This variant performs better than a strong baseline for RQL that does not use AIS losses. It is de

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyseyedsalehi2023approximate

Related papers

• Approximate Information State For Approximate Planning And Reinforcement Learning In Partially Observed Systems (2020)0.00
• Recurrent Networks, Hidden States And Beliefs In Partially Observable Environments (2022)0.00
• Deep Q-learning: Theoretical Insights From An Asymptotic Analysis (2020)10.35
• On Learning History Based Policies For Controlling Markov Decision Processes (2022)0.00
• Hidden Markov Model Estimation-based Q-learning For Partially Observable Markov Decision Process (2018)5.84
• A Finite-time Analysis Of Q-learning With Neural Network Function Approximation (2019)0.00
• Provably Efficient Reinforcement Learning In Partially Observable Dynamical Systems (2022)0.00
• Real-time Recurrent Reinforcement Learning (2023)2.26