Value Function Approximation in Zero-Sum Markov Games

Michail Lagoudakis, Ron Parr

This paper investigates value function approximation in the context of zero-sum Markov games, which can be viewed as a generalization of the Markov decision process (MDP) framework…