Explicit Manifold Representations for Value-Function Approximation in Reinforcement Learning.

William D. Smart

Published in: AI&M (2004)

Keyphrases

reinforcement learning
temporal difference
state space
temporal difference learning
function approximators
approximate dynamic programming
function approximation

reinforcement learning algorithms
markov games
state action
markov decision processes
low dimensional
manifold learning
learning algorithm

control problems
optimal control
high dimensional
model free
euclidean space
basis functions
robotic control

fixed point
policy iteration
learning process
multiple representations
fisher information
feature selection
machine learning