How Can Metaheuristics Help Software Engineers

Metaheuristics are approximate algorithms that can be used to face complex (e.g. NP-hard) problems of large dimensions, highly constrained, with mixed types of variables, and in general with complex search landscapes. Either in isolation or hybridized, evolutionary algorithms, ant colony optimization, particle swarm optimization, and traditional simulated annealing are powerful search procedures that an engineer can use to find high quality solutions to his/her problem. This keynote will be devoted to present the basic behavior of metaheuristics and to show how they can be applied to actual software engineering problems. We will discuss the main features of both, techniques and applications, to draw an efficient and accurate cross fertilization among them. Intelligent systems, operations research, computer science, and software engineering, all working together form a new field where many past difficult problems can be solved and where the present frontiers can be expanded by using new tools and research approaches.