Assumption-based pruning in conditional CSP

14Citations
Citations of this article
3Readers
Mendeley users who have this article in their library.
Get full text

Abstract

A conditional constraint satisfaction problem (CCSP) is a variant of the standard constraint satisfaction problem (CSP). CCSP's model problems where some of the variables and constraints may be conditionally inactive such that they do not participate in a solution. Recently, algorithms were introduced that use MAC at their core to solve CCSP. We extend MAC with a simple assumption-based reasoning. The resulting algorithm, Activity MAC (AMAC), is able to achieve significantly better pruning than existing methods. AMAC is shown to be more than two orders of magnitude more efficient than CondMAC on certain problem classes. Our algorithm is most naturally expressed using a variant of the CCSP representation that we refer to as Activity CSP (ACSP). ACSP introduces activity variablas which explicitly control the presence of other variables in the solution. Common aspects of CCSP, such as activity clustering and disjunction, are easily captured by ACSP and contribute to improved pruning by AMAC. © Springer-Verlag Berlin Heidelberg 2005.

Cite

CITATION STYLE

APA

Geller, F., & Veksler, M. (2005). Assumption-based pruning in conditional CSP. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 3709 LNCS, pp. 241–255). https://doi.org/10.1007/11564751_20

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free