102
click to vote
FLAIRS
2006
15 years 4 months ago
2006
We propose a new framework for aiding a reinforcement learner by allowing it to relocate, or move, to a state it selects so as to decrease the number of steps it needs to take in ...