Estimation and Approximation Bounds for Gradient-Based Reinforcement Learning.

Peter L. Bartlett, Jonathan Baxter

Browse the full COLT paper archive.