Efficient Inference in Markov Control Problems
This work addresses computational bottlenecks in reinforcement learning for researchers and practitioners, offering incremental improvements in efficiency for policy optimization methods.
The paper tackles the problem of inefficient marginal inference in Markov control problems by introducing a new exact inference algorithm that is more efficient than standard forward-backward recursions for finite horizons and extends to infinite horizons, providing a novel algorithm for policy gradients and Expectation Maximisation.
Markov control algorithms that perform smooth, non-greedy updates of the policy have been shown to be very general and versatile, with policy gradient and Expectation Maximisation algorithms being particularly popular. For these algorithms, marginal inference of the reward weighted trajectory distribution is required to perform policy updates. We discuss a new exact inference algorithm for these marginals in the finite horizon case that is more efficient than the standard approach based on classical forward-backward recursions. We also provide a principled extension to infinite horizon Markov Decision Problems that explicitly accounts for an infinite horizon. This extension provides a novel algorithm for both policy gradients and Expectation Maximisation in infinite horizon problems.