UAI
2008
13 years 6 months ago
2008
An efficient policy search algorithm should estimate the local gradient of the objective function, with respect to the policy parameters, from as few trials as possible. Whereas m...