ICML
2005
IEEE
15 years 10 days 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...