515 search results - page 56 / 103 » Approximating Markov Processes by Averaging |
225
click to vote
NIPS
2007
15 years 7 months ago
2007
Bayesian Reinforcement Learning has generated substantial interest recently, as it provides an elegant solution to the exploration-exploitation trade-off in reinforcement learning...
174
click to vote
TSP
2008
15 years 6 months ago
2008
Markov random fields are designed to represent structured dependencies among large collections of random variables, and are well-suited to capture the structure of real-world sign...
167
click to vote
GLOBECOM
2007
IEEE
16 years 13 days ago
2007
IEEE
— Adaptive transceivers play an important role in wireless communications and the design of MIMO systems. Therefore models that enable simulation of dynamic and time varying chan...
173
click to vote
ICML
2007
IEEE
16 years 7 months ago
2007
IEEE
This paper introduces a new approach to actionvalue function approximation by learning basis functions from a spectral decomposition of the state-action manifold. This paper exten...
155
click to vote
ICML
2006
IEEE
16 years 7 months ago
2006
IEEE
In the model-based policy search approach to reinforcement learning (RL), policies are found using a model (or "simulator") of the Markov decision process. However, for ...