81
Voted
ICML
2010
IEEE
15 years 11 hour ago
2010
IEEE
We consider approximate policy evaluation for finite state and action Markov decision processes (MDP) in the off-policy learning context and with the simulation-based least square...