GENET: a connectionist architecture for solving constraint satisfaction problems by iterative improvement

TLDR

Iterative improvement techniques solve large constraint satisfaction problems but can become trapped in local minima on highly constrained instances. This paper introduces GENET, a connectionist architecture for solving binary and general constraint satisfaction problems via iterative improvement. GENET uses a learning strategy to escape local minima, enabling it to iteratively improve solutions. Empirical results show that even when simulated on a single processor, GENET outperforms existing iterative improvement methods on hard instances of certain constraint satisfaction problems.

Abstract

New approaches to solving constraint satisfaction problems using iterative improvement techniques have been found to be successful on certain, very large problems such as the million queens. However, on highly constrained problems it is possible for these methods to get caught in local minima. In this paper we present GENET, a connectionist architecture for solving binary and general constraint satisfaction problems by iterative improvement. GENET incorporates a learning strategy to escape from local minima. Although GENET has been designed to be implemented on VLSI hardware, we present empirical evidence to show that even when simulated on a single processor GENET can outperform existing iterative improvement techniques on hard instances of certain constraint satisfaction problems.