ICML
2005
IEEE
14 years 11 months ago
2005
IEEE
Gaussian Process Temporal Difference (GPTD) learning offers a Bayesian solution to the policy evaluation problem of reinforcement learning. In this paper we extend the GPTD framew...