Nearly optimal stationary policies in negative dynamic programming.

Rolando Cavazos-Cadena Raúl Montes-de-Oca

Published in: Math. Methods Oper. Res. (1999)

Keyphrases

dynamic programming
stationary policies
markov decision processes
state space
optimal control
action sets
multistage
reinforcement learning
optimal policy
linear program
generative model
finite state
markov decision process