A number of methodological papers published during the last years testify that a need for a thorough revision of the research methodology is felt by the operations research community - see, for example, [Barr et al., J. Heuristics 1 (1995) 9-32; Eiben and Jelasity, Proceedings of the 2002 Congress on Evolutionary Computation (CEC'2002) 582-587; Hooker, J. Heuristics 1 (1995) 33-42; Rardin and Uzsoy, J. Heuristics 7 (2001) 261-304]. In particular, the performance evaluation of nondeterministic methods, including widely studied metaheuristics such as evolutionary computation and ant colony optimization, requires the definition of new experimental protocols. A careful and thorough analysis of the problem of evaluating metaheuristics reveals strong similarities between this problem and the problem of evaluating learning methods in the machine learning field. In this paper, we show that several conceptual tools commonly used in machine learning - such as, for example, the probabilistic notion of class of instances and the separation between the training and the testing datasets - fit naturally in the context of metaheuristics evaluation. Accordingly, we propose and discuss some principles inspired by the experimental practice in machine learning for guiding the performance evaluation of optimization algorithms. Among these principles, a clear separation between the instances that are used for tuning algorithms and those that are used in the actual evaluation is particularly important for a proper assessment.
@article{ITA_2006__40_2_353_0,
author = {Birattari, Mauro and Zlochin, Mark and Dorigo, Marco},
title = {Towards a theory of practice in metaheuristics design : a machine learning perspective},
journal = {RAIRO - Theoretical Informatics and Applications - Informatique Th\'eorique et Applications},
volume = {40},
year = {2006},
pages = {353-369},
doi = {10.1051/ita:2006009},
mrnumber = {2252644},
zbl = {1112.68109},
language = {en},
url = {http://dml.mathdoc.fr/item/ITA_2006__40_2_353_0}
}
Birattari, Mauro; Zlochin, Mark; Dorigo, Marco. Towards a theory of practice in metaheuristics design : a machine learning perspective. RAIRO - Theoretical Informatics and Applications - Informatique Théorique et Applications, Tome 40 (2006) pp. 353-369. doi : 10.1051/ita:2006009. http://gdmltest.u-ga.fr/item/ITA_2006__40_2_353_0/
[1] ,,, and, Designing and reporting computational experiments with heuristic methods. J. Heuristics 1 (1995) 9-32. | Zbl 0853.68154
[2] , The Problem of Tuning Metaheuristics, as Seen from a Machine Learning Perspective. Ph.D. thesis, Université Libre de Bruxelles, Brussels, Belgium (2004). | Zbl 1101.68748
[3] ,, and, A racing algorithm for configuring metaheuristics, in Proceedings of the Genetic and Evolutionary Computation Conference (GECCO-2002), edited by W.B. Langdon, et al. Morgan Kaufmann Publishers, San Francisco, CA (2002) 11-18.
[4] , and, Differential approximation algorithms for some combinatorial optimization problems. Theoret. Comput. Sci. 209 (1998) 107-122. | Zbl 0912.68061
[5] and, Ant colony optimization theory: A survey. Theoret. Comput. Sci. 344 (2005) 243-278. | Zbl pre02235590
[6] and, The Ant Colony Optimization meta-heuristic, in New Ideas in Optimization, edited by D. Corne, M. Dorigo and F. Glover. McGraw Hill, London, UK (1999) 11-32.
[7] , and, Ant algorithms for discrete optimization. Artificial Life 5 (1999) 137-172.
[8] , and, Ant System: Optimization by a colony of cooperating agents. IEEE Trans. Systems, Man, and Cybernetics - Part B 26 (1996) 29-41.
[9] and, Ant Colony Optimization. MIT Press, Cambridge, MA (2004). | Zbl 1092.90066
[10] and, A Critical Note on Experimental Research Methodology in EC, in Proceedings of the 2002 Congress on Evolutionary Computation (CEC'2002), Piscataway, NJ, IEEE Press (2002) 582-587.
[11] , and, Artificial Intelligence Through Simulated Evolution. John Wiley & Sons, New York, NY (1966). | Zbl 0148.40701
[12] and, Computers and Intractability: A Guide to the Theory of -Completeness. Freeman, San Francisco, CA (1979). | MR 519066 | Zbl 0411.68039
[13] , Tabu search - part I. ORSA J. Comput. 1 (1989) 190-206. | Zbl 0753.90054
[14] , Tabu search - part II. ORSA J. Comput. 2 (1990) 4-32. | Zbl 0771.90084
[15] , Genetic Algorithms in Search, Optimization, and Machine Learning. Addison-Wesley, Reading, MA (1989). | Zbl 0721.68056
[16] and, Z-approximations. J. Algorithms 41 (2001) 429-442. | Zbl 1014.68222
[17] , Adaptation in Natural and Artificial Systems. University of Michigan Press, Ann Arbor, MI (1975). | MR 441393 | Zbl 0317.68006
[18] , Testing heuristics: we have it all wrong. J. Heuristics 1 (1995) 33-42. | Zbl 0853.68155
[19] , A theoretician's guide to the experimental analysis of algorithms, in Data structures, near neighbor searches, and methodology: 5th and 6th DIMACS implementation challenges. American Mathematical Society, Providence, RI (2002) 215-250. | Zbl 1103.68997
[20] , The Origins of Order. Self-Organization and Selection in Evolution. Oxford University Press, Oxford, UK (1993).
[21] , and, Optimization by simulated annealing. Science 220 (1983) 671-680.
[22] , and, Adapting self-adaptive parameters in evolutionary algorithms. Appl. Intell. 15 (2001) 171-180. | Zbl 0992.68231
[23] , and, Iterated local search, in Handbook of Metaheuristics. International Series in Operations Research & Management Science, edited by F. Glover and G. Kochenberger. Kluwer Academic Publishers, Norwell, MA 57 (2002) 321-353. | Zbl 1116.90412
[24] , Introduction to Linear and Nonlinear Programming. Addison-Wesley Publishing Company, Reading, MA (1973). | Zbl 0297.90044
[25] and, Hoeffding races: Accelerating model selection search for classification and function approximation, in Advances in Neural Information Processing Systems, edited by J.D. Cowan, G. Tesauro and J. Alspector. Morgan Kaufmann Publishers, San Francisco, CA 6 (1994) 59-66.
[26] , Toward an experimental method for algorithm simulation. INFORMS J. Comput. 2 (1996) 1-15. | Zbl 0854.68038
[27] and, How to Solve it: Modern Heuristics. Springer-Verlag, Berlin, Germany (2000). | MR 1730907 | Zbl 0943.90002
[28] and, Efficient algorithms for minimizing cross validation error, in International Conference on Machine Learning. Morgan Kaufmann Publishers, San Francisco, CA (1994) 190-198.
[29] ,, and, Simple procedures for selecting the best simulated system when the number of alternatives is large. Oper. Res. 49 (2001) 950-963.
[30] and, Experimental evaluation of heuristic optimization algorithms: A tutorial. J. Heuristics 7 (2001) 261-304. | Zbl 0972.68634
[31] , Evolutionsstrategie - Optimierung technischer Systeme nach Prinzipien der biologischen Information. Fromman Verlag, Freiburg, Germany (1973).
[32] , Numerical Optimization of Computer Models. John Wiley & Sons, Chichester, UK (1981). | Zbl 0451.65043
[33] , Software Engineering. Addison Wesley, Harlow, UK, sixth edition (2001). | Zbl 0557.68006
[34] , Self-adaptive exploration in evolutionary search. Technical Report IRINI-2001-05, Institut für Neuroinformatik, Ruhr-Universität Bochum, Bochum, Germany (2001).
[35] and, No free lunch theorems for optimization. IEEE Transactions on Evolutionary Computation 1 (1997) 67-82.
[36] , Measuring the quality of approximate solutions to zero-one programming problems. Math. Oper. Res. 6 (1981) 319-332. | Zbl 0538.90065
[37] ,, and, Model-based search for combinatorial optimization: A critical survey. Ann. Oper. Res. 131 (2004) 375-395. | Zbl 1067.90162
[38] and, Model based search for combinatorial optimization: A comparative study, in Parallel Problem Solving from Nature - PPSN VII, edited by M. Guervós, J.J. et al. Springer Verlag, Berlin, Germany (2002) 651-661.