Abstract : As Lee and Leung have previously shown this year, weighted constraint satisfaction problems can benefit from the introduction of global cost functions, leading to a new Cost Function Programming paradigm. In this paper, we explore the possibility of decomposing global cost functions in such a way that enforcing soft local consistencies on the decomposition achieves the same level of consistency on the original global cost function. We give conditions under which directional and virtual arc consistency offer such guarantees. We conclude by experiments on decomposable cost functions showing that decompositions may be very useful to easily integrate efficient global cost functions in solvers.
Type de document :
Communication dans un congrès
AAAI'12: Conference on Artificial Intelligence, Toronto, Ontario, Canada. pp.7, 2012
https://hal-lirmm.ccsd.cnrs.fr/lirmm-00748187
Contributeur : Joël Quinqueton
<>
Soumis le : vendredi 9 novembre 2012 - 17:26:41
Dernière modification le : jeudi 28 juin 2018 - 14:35:18
Document(s) archivé(s) le : dimanche 10 février 2013 - 02:55:08
David Allouche, Christian Bessière, Patrice Boizumault, Simon De Givry, Patricia Gutierrez, et al.. Filtering Decomposable Global Cost Functions. AAAI'12: Conference on Artificial Intelligence, Toronto, Ontario, Canada. pp.7, 2012. 〈lirmm-00748187〉
