The whole greater than the sum of its parts: maximizing the strengths of heuristic set covering algorithms

We describe a general procedure that maximizes the strengths of selected heuristic optimization algorithms to produce solutions that are often better than what each algorithm could produce individually. We illustrate our procedure on the generalized set covering problem. By combining several heuristic algorithms, we obtain optimal solutions in many instances. The algorithms used in this way are shown to be effective also in solving the classical set covering problem.