Correlation Heuristics for Constraint Programming

TL;DR

This paper introduces correlation-based heuristics for guiding search in constraint programming, using variable correlations measured during constraint propagation, which are competitive and often faster than existing heuristics.

Contribution

It proposes novel correlation heuristics, crbs-sum and crbs-max, for variable selection in constraint programming, demonstrating their effectiveness through extensive benchmarking.

Findings

Correlation heuristics are competitive with established methods.

Correlation heuristics often achieve the fastest search times.

Experiments show the effectiveness of correlation-based guidance.

Abstract

Effective general-purpose search strategies are an important component in Constraint Programming. We introduce a new idea, namely, using correlations between variables to guide search. Variable correlations are measured and maintained by using domain changes during constraint propagation. We propose two variable heuristics based on the correlation matrix, crbs-sum and crbs-max. We evaluate our correlation heuristics with well known heuristics, namely, dom/wdeg, impact-based search and activity-based search. Experiments on a large set of benchmarks show that our correlation heuristics are competitive with the other heuristics, and can be the fastest on many series.