The divergence of reinforcement learning algorithms with value-iteration and function approximation.

Michael Fairbank, Eduardo Alonso

Browse the full IJCNN paper archive.