Appendices A Reinforcement Learning using Matrix Estimation: The Pseudo Code

May-30-2025, 01:17:07 GMT–Neural Information Processing Systems 

For the mean error, we use the average of the (absolute) difference over this grid.