Robust Policy Computation in Reward-Uncertain MDPs Using Nondominated Policies

AAAI, 2010.

Cited by: 50|Views11
EI

Abstract:

The precise specification of reward functions for Markov de- cision processes (MDPs) is often extremely difficult, motiv at- ing research into both reward elicitation and the robust sol u- tion of MDPs with imprecisely specified reward (IRMDPs). We develop new techniques for the robust optimization of IR- MDPs, using the minimax regret de...More

Code:

Data:

Your rating :
0

 

Tags
Comments