Exploiting Belief State Structure in Graph Search

作者: Stuart Russell , Jason Wolfe

DOI:

关键词:

摘要: It is well-known that eliminating repeated states essential for efficient search of state-space AND-OR graphs. The same technique has been found useful searching beliefstate graphs, which arise in nondeterministic partially observable planning problems and games. Whereas physical are viewed by algorithms as atomic admit only equality tests, belief states, sets possible have additional structure: one state can subsume or be subsumed another. This paper presents new exploit this property to achieve substantial speedups. demonstrated on Kriegspiel checkmate a vacuum world domain.

参考文章(13)
Makoto Sakuta, Hiroyuki Iida, SOLVING KRIEGSPIEL-LIKE PROBLEMS: EXPLOITING A TRANSPOSITION TABLE ICGA Journal. ,vol. 23, pp. 218- 229 ,(2000) , 10.3233/ICG-2000-23403
James Kurien, David E. Smith, P. Pandurang Nayak, Fragment-based conformant planning international conference on artificial intelligence planning systems. pp. 153- 162 ,(2002)
Stuart Russell, Jason Wolfe, Efficient belief-state AND-OR search, with application to Kriegspiel international joint conference on artificial intelligence. pp. 278- 285 ,(2005)
Jorg Hoffmann, Jana Koehler, A new Method to Index and Query Sets international joint conference on artificial intelligence. pp. 462- 467 ,(1999)
L.V. Allis, Searching for solutions in games and artificial intelligence Ph. D. Thesis, University of Limburg. ,(1994)
D.M. Breuker, Memory versus search in games Universiteit Maastricht. ,(1998)
Alessandro Cimatti, Marco Roveri, Piergiorgio Bertoli, Paolo Traverso, Planning in nondeterministic domains under partial observability via symbolic model checking international joint conference on artificial intelligence. pp. 473- 478 ,(2001)
Murray Campbell, The graph-history interaction: on ignoring position history acm annual conference on range of computing. pp. 278- 280 ,(1985) , 10.1145/320435.320516
Robert C. Schrag, Roberto J. Bayardo, Using CSP look-back techniques to solve real-world SAT instances national conference on artificial intelligence. pp. 203- 208 ,(1997)