Approximate receding horizon approach for Markov decision processes: average reward case (Q1414220): Difference between revisions

The authors consider an approximation scheme for solving Markov decision processes (MDPs) with countable state space, finite action space, and bounded rewards that uses an approximate solution of a fixed finite-horizon sub-MDP of a given infinite-horizon MDP to create a stationary policy, which they call ''approximate receding horizon control''. They analyze the performance of the approximate receding horizon control in some conditions, study two examples, also provide a simple proof on the policy improvement for countable state space, and discuss practical implementations of these schemes via simulation.