ℓ1-Penalized Projected Bellman Residual.

Matthieu Geist Bruno Scherrer

Published in: EWRL (2011)

Keyphrases

bellman residual
least squares
policy iteration
sample path
policy evaluation
asymptotic analysis
markov decision processes
optimization criterion
hybrid algorithms
loss function
evaluation function
semi parametric
neural network
maximum likelihood
dynamic programming
optical flow
machine learning