AAAI Publications, Twenty-Sixth International Conference on Automated Planning and Scheduling

Font Size: 
Efficient Representation of Pattern Databases Using Acyclic Random Hypergraphs
Mehdi Sadeqi, Howard J. Hamilton

Last modified: 2016-03-30

Abstract


A popular way to create domain-independent heuristic functions is by using abstraction, where an abstract (coarse) version of the state space is derived from the original state space. An abstraction-based heuristic is usually implemented using a pattern database, a lookup table of (abstract state, heuristic value) pairs. Efficient representation and compression of pattern databases has been the topic of substantial ongoing research. In this paper, we present a novel domain-independent algorithm for this purpose using acyclic r-partite random hypergraphs. The theoretical and experimental results show that our proposed algorithm provides a consistent representation that works well across planning problem domains and provides a good trade-off between space usage and lookup time. Thus, it is suitable to be a standard efficient representation for pattern databases and a benchmark method for other pattern database representation/compression methods.

Full Text: PDF