"Reinforcement learning with Gaussian processes."

Yaakov Engel, Shie Mannor, Ron Meir (2005)
a service of Schloss Dagstuhl - Leibniz Center for Informatics