Applying marginal MAP search to probabilistic conformant planning: initial results
Junkyu Lee, Rina Dechter, Radu Marinescu · National Conference on Artificial Intelligence · 2014
In this position paper, we present our current progress in applying marginal MAP algorithms for solving the conformant planning problems. Conformant planning problem is formulated as probabilistic inference in graphical models compiled from relational PPDDL domains. The translation from PPDDL into Dynamic Bayesian Network is developed by mapping the SAT encoding of the ground PPDDL into factored representation. We experimented with recently developed AND/OR branch and bound search algorithms for marginal MAP over instances from the international planning competition domains, and we show that several domains were solved efficiently.