Malte Helmert, Patrik Haslum, Jörg Hoffmann
Many AI problems can be recast as finding an optimal path in a discrete state space. An abstraction defines an admissible heuristic function as the distances in a smaller state space where arbitrary sets of states are "aggregated" into single states. A special case are pattern database (PDB) heuristics, which aggregate states iff they agree on the state variables inside the pattern. Explicit-state abstraction is more flexible, explicitly aggregating selected pairs of states in a process that interleaves composition of abstractions with abstraction of the composites. The increased flexibility gains expressive power: sometimes, the real cost function can be represented concisely as an explicit-state abstraction, but not as a PDB. Explicit-state abstraction has been applied to planning and model checking, with highly promising empirical results.
Subjects: 15.7 Search; 1.11 Planning
Submitted: Apr 15, 2008