437 search results - page 11 / 88 » Policy Gradient Critics |
109
click to vote
ICML
2009
IEEE
16 years 3 months ago
2009
IEEE
In this paper we introduce the first algorithms for efficiently learning a simulation policy for Monte-Carlo search. Our main idea is to optimise the balance of a simulation polic...
122
Voted
JMLR
2006
2006
Geometric Variance Reduction in Markov Chains: Application to Value Function and Gradient Estimation
15 years 2 months ago
We study a sequential variance reduction technique for Monte Carlo estimation of functionals in Markov Chains. The method is based on designing sequential control variates using s...
142
Voted
EWRL
2008
15 years 4 months ago
2008
Policy Learning approaches are among the best suited methods for high-dimensional, continuous control systems such as anthropomorphic robot arms and humanoid robots. In this paper,...
122
click to vote
MICRO
2005
IEEE
15 years 8 months ago
2005
IEEE
Clustered machines partition hardware resources to circumvent the cycle time penalties incurred by large, monolithic structures. This partitioning introduces a long inter-cluster ...
128
click to vote
NIPS
2003
15 years 4 months ago
2003
We describe a new approximation algorithm for solving partially observable MDPs. Our bounded policy iteration approach searches through the space of bounded-size, stochastic fini...