Weak Convergence Properties of Constrained Emphatic Temporal-difference Learning with Constant and Slowly Diminishing Stepsize.

Published in: CoRR (2015)

Keyphrases

temporal difference learning
temporal difference
step size
convergence rate
faster convergence
convergence speed
quasi newton
function approximation
reinforcement learning
evaluation function
fixed point
model free
game playing
reinforcement learning algorithms
monte carlo
linear programming
genetic algorithm