Conditional Dominance Analysis for Classical Planning

Anna Wilhelm, Alvaro Torralba

Track: Long Paper (9 pages including references)

Previous Publication: Yes, the submission has already been published or acccepted at another conference.

Keywords: Dominance Analysis, Optimal Planning

Abstract:

Dominance analysis methods compare pairs of states in a planning task to prove that one is at least as close to the goal than other. Existing methods compute fact-dominance relations, which identify facts that are at least as good than others in any situation. However, this is only possible when a fact is as good as another in every single possible context.

We introduce a new notion of conditional dominance, which can identify that a fact dominates another under certain conditions. We extend previous methods to compute dominance by taking into account a set of ``contexts'' in order to find coarser dominance relations. We propose several strategies to find relevant contexts automatically and show that even with one single condition, one can achieve significant pruning in certain domains.

Submission Number: 10