Iterated hyperplane search for the budgeted maximum coverage problem

Abstract

We present an iterated hyperplane search approach for the budgeted maximum coverage problem. Our algorithm relies on the idea of searching on specific areas identified by cardinality-constrained hyperplanes. It combines three complementary procedures: a tabu search procedure to identify a promising hyperplane, a hyperplane search procedure to examine candidate solutions on cardinality-constrained hyperplanes and a dedicated perturbation procedure to ensure the diversification of the search. We show the competitiveness of the algorithm on 30 benchmark instances and present experiments to study the key components of the algorithm.