Il y a 1 fichier(s) dont l'accès est interdit, à la demande de l'auteur.
|Titre||The Problem of Tuning Metaheuristics as seen from a Machine Learning Perspective|
|Département||F517 - Faculté des sciences appliquées - Informatique|
|Intitulé du diplôme||Doctorat en sciences appliquées|
|Date de défense||2004-12-20|
Bersini, Hugues (Membre du jury/Committee Member)
Bonarini, Andrea (Membre du jury/Committee Member)
Bontempi, Gianluca (Membre du jury/Committee Member)
Decaestecker, Christine (Membre du jury/Committee Member)
Van Ham, Philippe (Président du jury/Committee Chair)
Dorigo, Marco (Promoteur/Director)
|Mots-clés||Combinatorial optimization, Racing algorithms, Tuning, Metaheuristics|
A metaheuristic is a generic algorithmic template that, once properly instantiated, can be used for finding high quality solutions of combinatorial optimization problems. For obtaining a fully functioning algorithm, a metaheuristic needs to be configured: typically some modules need to be instantiated and some parameters need to be tuned. For the sake of precision, we use the expression parametric tuning for referring to the tuning of numerical parameters, either continuous or discrete but in any case ordinal. On the other hand, we use the expression structural tuning for referring to the problem of defining which modules should be included and, in general, to the problem of tuning parameters that are either boolean or categorical. Finally, with tuning we refer to the composite structural and parametric tuning.
Tuning metaheuristics is a very sensitive issue both in practical applications and in academic studies. Nevertheless, a precise definition of the tuning problem is missing in the literature. In this thesis, we argue that the problem of tuning a metaheuristic can be profitably described and solved as a machine learning problem.
Indeed, looking at the problem of tuning metaheuristics from a machine learning perspective, we are in the position of giving a formal statement of the tuning problem and to propose an algorithm, called F-Race, for tackling the problem itself. Moreover, always from this standpoint, we are able to highlight and discuss some catches and faults in the current research methodology in the metaheuristics field, and to propose some guidelines.
The thesis contains experimental results on the use of F-Race and some examples of practical applications. Among others, we present a feasibility study carried out by the German-based software company SAP, that concerned the possible use of F-Race for tuning a commercial computer program for vehicle routing and scheduling problems. Moreover, we discuss the successful use of F-Race for tuning the best performing algorithm submitted to the International Timetabling Competition organized in 2003 by the Metaheuristics Network and sponsored by PATAT, the international series of conferences on the Practice and Theory of Automated Timetabling.