Generalizing Constraint Models in Constraint Acquisition

TL;DR

This paper introduces GenCon, a method that uses statistical learning to generalize constraint models across different problem instances, enabling flexible and interpretable constraint specifications.

Contribution

The paper presents GenCon, a novel approach that learns parameterized constraint models capable of generalizing across multiple problem instances using classifiers and decision rules.

Findings

High accuracy in predicting constraints across instances

Robustness to noise in input data

Effective generation of ground constraints on demand

Abstract

Constraint Acquisition (CA) aims to widen the use of constraint programming by assisting users in the modeling process. However, most CA methods suffer from a significant drawback: they learn a single set of individual constraints for a specific problem instance, but cannot generalize these constraints to the parameterized constraint specifications of the problem. In this paper, we address this limitation by proposing GenCon, a novel approach to learn parameterized constraint models capable of modeling varying instances of the same problem. To achieve this generalization, we make use of statistical learning techniques at the level of individual constraints. Specifically, we propose to train a classifier to predict, for any possible constraint and parameterization, whether the constraint belongs to the problem. We then show how, for some classes of classifiers, we can extract decision rules to construct interpretable constraint specifications. This enables the generation of ground constraints for any parameter instantiation. Additionally, we present a generate-and-test approach that can be used with any classifier, to generate the ground constraints on the fly. Our empirical results demonstrate that our approach achieves high accuracy and is robust to noise in the input instances.