QL2, a simple reinforcement learning scheme for two-player zero-sum Markov games.

Benot Frnay, Marco Saerens

Browse the full ESANN paper archive.