An immune-inspired algorithm for the set cover problem

Abstract

This paper introduces a novel parallel immune-inspired algorithm based on recent developments in the understanding of the germinal centre reaction in the immune system. Artificial immune systems are relatively new randomised search heuristics and work on parallelising them is still in its infancy. We compare our algorithm with a parallel implementation of a simple multi-objective evolutionary algorithm on benchmark instances of the set cover problem taken from the OR-library. We show that our algorithm finds feasible solutions faster than the evolutionary algorithm using less parameters and communication effort.