Abstract
Stochastic search is a key mechanism underlying many metaheuristics. The chapter starts with the presentation of a general framework algorithm in the form of a stochastic search process that contains a large variety of familiar metaheuristic techniques as special cases. Based on this unified view, questions concerning convergence and runtime are discussed on the level of a theoretical analysis. Concrete examples from diverse metaheuristic fields are given. In connection with runtime results, important topics as instance difficulty, phase transitions, parameter choice, No-Free-Lunch theorems, or fitness landscape analysis are addressed. Furthermore, a short sketch of the theory of black-box optimization is given, and generalizations of results to stochastic search under noise are outlined.
Notes
- 1.
Since g and h do not depend on the iteration counter t, the Markov process is homogeneous. Dependence on t can easily be modeled by adding t as a component to the memory M t .
- 2.
Elitism as a mechanism ensuring convergence of a GA has already been analyzed in [39], which appears to be the first paper on GA convergence.
- 3.
To ask, say, for the expected time until first hitting an optimal solution without being sure that the optimum will be reached, is as meaningless as to ask “How much training time would it take in the average for a randomly selected person to win an olympic gold medal?” Also by being content with an approximate solution of a certain minimum quality (call it the “silver medal”) instead of the optimal solution, one does not escape this difficulty.
- 4.
The relevance of this distribution in the field of stochastic search is also underlined by the fact that one of the oldest general-purpose stochastic search techniques, namely SA, approximates at each fixed temperature level T the corresponding Boltzmann distribution.
- 5.
There seem to be close relations between NFL theorems and the well-known philosophical induction problem that plays also a role in AI approaches to inductive reasoning. Suppose that the function evaluation of \(f(x)\) in the solution \(x \in S\) is not done by an algorithmic computation, but rather by the observation of some real-world system (say, x is a control vector for a chemical plant, and \(f(x)\) is the observed value of an outcome variable). Then the “NFL insight” that there is no logical argument why the observation of \(f(x_1),\ldots,f(x_{t-1})\) for some sample solutions \(x_1,\ldots,x_{t-1}\) should provide any information on \(f(x_t)\) for the sample solution \(x_t \notin \{x_1,\ldots,x_{t-1}\}\), basically amounts to the intriguing claim by David Hume that we do not have any logical justification for the step called “inductive conclusion,” although this step is indispensable in science as in everyday life.
- 6.
- 7.
For approaches using “white-box” mathematical programming techniques such as the Integer L-Shaped Method, see, e.g., [23].
References
Achlioptas, D., Naor, A., Peres, Y.: Rigorous location of phase transitions in hard optimization problems. Nature 435, 759–764 (2005)
A simulated annealing algorithm with constant temperature for discrete stochastic optimization. Manage. Sci. 45, 748–764 (1999)
Barbosa, V.C., Ferreira, R.G.: On the phase transitions of graph coloring and independent sets. Physika A 343, 401–423 (2004)
Bianchi, L., Dorigo, M., Gambardella, L.M., Gutjahr, W.J.: A survey on metaheuristics for stochastic combinatorial optimization. To appear In: Natural Computing 8, 239–287 (2009)
Birattari, M., Balaprakash, P., Dorigo, M.: The ACO/FRACE algorithm for combinatorial optimization under uncertainty. In: Doerner K. et al. (ed.) Metaheuristics – Progress in Complex Systems Optimization, Springer: Berlin, Germany (2006)
Borenstein, Y., Poli, R.: Information Perspective of Optimization. Proceedings of 9th Conference on Parallel Problem Solving from Nature, Springer LNCS, vol. 4193, pp. 102–111. Berlin Heidelberg (2006)
Borenstein, Y. Poli, R.: Structure and Metaheuristics. Proceedings of Genetic and Evolutionary Computation Conference ’06, pp. 1087–1093 (2006)
Borisovsky, P.A., Eremeev, A.V.: A study on performance of the (1+1)-evolutionary algorithm. In: Proceedings of Foundations of Genetic Algorithms, vol. 7, pp. 271–287. Morgan Kaufmann, San Francisco (2003)
Cheesman, P., Kenafsky, B., Taylor, W.M.: Where the really hard problems are. In: Morgan Kaufmann (ed.) Proceedings of IJCAI ’91, pp. 331–337 (1991)
Droste, S., Jansen, T., Wegener, I.: Perhaps not a free lunch but at least a free appetizer. Proceedings of Genetic and Evolutionary Computation Conference ’99, Orlando, FL, USA pp. 833–839 (1999)
Droste, S., Jansen, T., Wegener, I.: On the analysis of the (1+1) evolutionary algorithm. Theor. Comput. Sci. 276, 51–81 (2002)
Droste, S., Jansen, T., Wegener, I.: Upper and lower bounds for randomized search heuristics in black-box optimization. Theory Comput. Syst. 39 525–544 (2006)
Dorigo, M., Maniezzo, V., Colorni, A.: Ant System: optimization by a colony of cooperating agents. IEEE Trans. Syst, Man Cybern 26, 1–13 (1996)
English, T.: Optimization is easy and learning is hard in the typical function. Proceedings of Congress in Evolutionary Computation ’00, La Jolla, USA pp. 924–931 (2000)
English, T.: On the structure of sequential search: beyond “no free lunch”. Proc. EvoCOP ’04, Springer LNCS, Coimbra, Portugal 3004, 95–103 (2004)
Eremeev, A.V., Reeves, C.R.: On confidence intervals for the number of local optima. Applications of Evolutionary Computing, Springer LNCS, Berlin Heidelberg vol. 2611, pp. 224–235 (2003)
Ferreira, F.F., Fontanari, J.F.: Probabilistic analysis of the number partitioning problem. J. Phys. A: Math. Gen. 31, 3417–3428 (1998)
Fischetti, M., Lodi, A.: Local branching. Mathematical Programming Ser. B 98, 23–47 (2003)
Friedrich, T., He, J., Hebbinghaus, N., Neumann, F., Witt, C.: Approximating covering problems by randomized search heuristics using multi-objective models. Proceedings of 9th Annual Conference on Genetic and Evolutionary Computation, pp. 797–804 (2007)
Garnier, J., Kalel, L., Schoenauer, M.: Rigorous hitting times for binary mutations. Evol. Comput. 7, 45–68 (1999)
Garnier, J., Kallel, L.: Efficiency of local search with multiple local optima. SIAM J. Discrete Math. 15, 122–141 (2002)
Gelfand, S.B., Mitter, S.K.: Simulated annealing with noisy or imprecise measurements. J. Optim. Theor. Appl. 69, 49–62 (1989)
Gendreau, M., Laporte, G., Seguin, R.: An exact algorithm for the vehicle routing problem with stochastic demands and customers. Transport. Sci. 29, 143–155 (1995)
Gent, I.P., Walsh, T.: Analysis of heuristics for number partitioning. Comput. Intell. 14, 430–450 (1998)
Giel, O., Wegener, I.: Evolutionary algorithms and the maximum matching problem. Proceedings of 20th Annual Symposium on Theoretical Aspects of Computer Science, Seattle, Washington, USA pp. 415–426 (2003)
González, C., Lozano, J.A., Larrañaga, P.: Analyzing the PBIL algorithm by means of discrete dynamical systems. Complex Syst. 11, 1–15 (1997)
González, C., Lozano, J.A., Larrañaga, P.: Mathematical modelling of discrete estimation of distribution algorithms. In: Larrañaga et al. (eds.) Estimation of Distribution Algorithms, A New Tool for Evolutionary Computation, Kluwer, Dordrecht 147–163 (2002)
Gutjahr, W.J.: A graph–based ant system and its convergence. Future Gen. Comput. Syst. 16, 873–888 (2000)
Gutjahr, W.J.: ACO algorithms with guaranteed convergence to the optimal solution. In. Process. Lett. 82, 145–153 (2002)
Gutjahr, W.J.: A converging ACO algorithm for stochastic combinatorial optimization. Proceedings of 2nd Symposium on Stochastic Algorithms, Foundations and Applications, Springer LNCS, Berlin Heidelberg vol. 2827, pp. 10–25 (2003)
Gutjahr, W.J.: S-ACO: An ant-based approach to combinatorial optimization under uncertainty. Proceedings of 4nd Int. Workshop on Ant Colony Optimization and Swarm Intelligence, Springer LNCS, Berlin Heidelberg New York vol. 3172, pp. 238–249 (2004)
Gutjahr, W.J.: On the finite-time dynamics of ant colony optimization. Methodol. Comput. Appl. Probability 8, 105–133 (2006)
Gutjahr, W.J.: Mathematical runtime analysis of ACO algorithms: survey on an emerging issue. Swarm Intell. 1, 59–79 (2007)
Gutjahr, W.J.: First steps to the runtime complexity analysis of ant colony optimization. Comput. Oper. Res. 35, 2711–2727 (2008)
Gutjahr, W.J., Katzensteiner, S., Reiter, P.: A VNS algorithm for noisy problems and its application to project portfolio analysis. Proceedings of SAGA 2007 (Stochastic Algorithms: Foundations and Applications), Springer LNCS, Berlin Heidelberg vol. 4665, pp. 93–104 (2007)
Gutjahr, W.J., Pflug, G.: Simulated annealing for noisy cost functions. J. Global Optim. 8, 1–13 (1996)
Gutjahr, W.J., Sebastiani, G.: Runtime analysis of ant colony optimization with best-so-far reinforcement. Methodology and Computing in Applied Probability 10, 409–433 (2008)
Hajek, B.: Cooling schedules for optimal annealing. Math. of Operat. Res. 13, 311–329 (1988)
Hartl, R.F.: A global convergence proof for a class of genetic algorithms. Technical Report, University of Vienna (1990)
Hartmann, A.K., Barthel, W., Weigt, M.: Phase transition and finite-size scaling in the vertex-cover problem. Comput. Phys. Commn. 169, 234–237 (2005)
He, J., Yao, X.: Drift analysis and average time complexity of evolutionary algorithms. Artifi. Intell. 127, 57–85 (2003)
He, J., Yao, X.: Towards an analytic framework for analysing the computation time of evolutionary algorithms. Artif. Intell. 145, 59–97 (2003)
He, J., Yao, X.: A study of drift analysis for estimating computation time of evolutionary algorithms. Nat. Comput. 3, 21–35 (2004)
He, J., Yu, X.: Conditions for the convergence of evolutionary algorithms. J. Syst. Arch. 47, 601–612 (2001)
Herroelen, W., De Reyck, B.: Phase transitions in project scheduling. J. Oper. Res. Soc. 50, 148–156 (1999)
Igel, C., Toussaint, M.: On classes of functions for which no free lunch results hold. Inf. Process. Lett. 86, 317–321 (2003)
Igel, C., Toussaint, M.: A no-free-lunch theorem for non-uniform distributions of target functions. J. Math. Model. Algorithms 3, 313–322 (2004)
Jacobson, S.H., Yücesan, E.: Analyzing the performance of generalized hill climbing algorithms. J. Heuristics 10, 387–405 (2004)
Jansen, T., Wegener, I.: On the analysis of a dynamic evolutionary algorithm. J. Discrete Algorithms 4, 181–199 (2006)
Jin, Y., Branke, J.: Evolutionary optimization in uncertain environments – a survey. IEEE Trans. Evol. Comput. 9, 303–317 (2005).
Kennedy, J., Eberhart, R.C.: A discrete binary version of the particle swarm algorithm. Proceedings of the World Multiconference on Systemics, Cybernetics and Informatics, Orlando, FL, USA pp. 4104–4109 (1997)
Koehler, G.J.: Conditions that obviate the no-free-lunch theorems for optimization. Informs J. Comput. 19, 273–279 (2007)
Ladret, V.: Asymptotic hitting time for a simple evolutionary model of protein folding. J. Appl. Probability 42, 39–51 (2005)
Margolin, L.: On the convergence of the cross-entropy method. Ann. Oper. Res. 134, 201–214 (2005)
Martin, O.C., Monasson, R., Zecchina, R.: Statistical mechanics methods and phase transitions in optimization problems. Theor. Compu. Sci. 265, 3–67 (2001)
Merelo, J.-J., Cotta, C.: Building bridges: the role of subfields in metaheuristics. SIGEVOlution 1(4), 9–15 (2006)
Mertens, S.: A physicist’s approach to number partitioning. Theor. Comput. Sci. 265, 79–108 (2001)
Merz, P., Freisleben, B.: Fitness landscape analysis and memetic algorithms for the quadratic assignment problem. IEEE Trans. Evol Comput 4, 337–352 (2000)
Monasson, R.: Introduction to phase transitions in random optimization problems. Technical Report, Laboratoire de Physique Theorique de l’ENS, Paris (2007)
Neumann, F., Wegener, I.: Randomized local search, evolutionary algorithms, and the minimum spanning tree problem. Theor Comp. Sci. 378, 32–40 (2007).
Neumann, F., Witt, C.: Runtime analysis of a simple ant colony optimization algorithm. Proceedings of ISAAC ’06, Springer LNCS, Berlin Heidelberg vol. 4288, pp. 618–627 (2006)
Neumann, F., Witt, C.: Ant colony optimization and the minimum spanning tree problem. Proceedings of LION ’08, Learning and Intelligent Optimization, Theoretical Computer Science 411, 2406–2413 (2010)
Norman, F.: Markov Processes and Learning Models. Academic Press, New York (1972)
Oliveto, P.S., He, J., Yao, X.: Time complexity of evolutionary algorithms for combinatorial optimization: a decade of results. Int. J. Automat. and Comput. 4, 281–293 (2007)
Oliveto, P.S., He, J., Yao, X.: Evolutionary algorithms and the vertex cover problem. Proceedings of the Congress on Evolutionary Computation CEC ’07, Singapore pp. 1870–1877 (2007)
Purkayastha, P., Baras, J.S.: Convergence results for ant routing algorithms via stochastic approximation and optimization. Proceedings of 46th IEEE Conference on Decision and Control, pp. 340–345 (2007)
Reidys C.M., Stadler, P.F.: Combinatorial landscapes. SIAM Rev. 44, 3–54 (2002)
Rudolph, G.: Convergence Analysis of canonical genetic algorithms. IEEE Trans. Neural. Netw. 5, 96–101 (1994)
Sasaki, G.H., Hajek, B.: The time complexity of maximum matching by simulated annealing. J. ACM 35, 67–89 (1988)
Scharnow, J., Tinnefeld, K., Wegener, I.: Fitness landscapes based on sorting and shortest path problems. Proceedings of 7th Conference on Parallel Problem Solving from Nature, 54–63 (2002)
Schiavinotto, T., Stûtzle, T.: A review of metrics on permutations for search landscape analysis. Comput. Oper. Res. 34, 3143–3153 (2007)
Sebastiani, G., Torrisi, G.L.: An extended ant colony algorithm and its convergence analysis. Methodol. Comput. Appl. Probability 7, 249–263 (2005)
Storch, T.: How randomized search heuristics find maximum cliques in planar graphs. Proceedings of 8th Annual Conference on Genetic and Evolutionary Computation, Seattle, Washington, USA pp. 567–574 (2006)
Stützle, T., Hoos, H.H.: MAX-MIN Ant System. Future Gen. Comput. Sys. 16, 889–914 (2000)
Stützle, T., Dorigo, M.: A short convergence proof for a class of ACO algorithms. IEEE Trans. Evol. Comput. 6, 358–365 (2002)
Sudholt, D., Witt, C.: Runtime analysis of binary PSO. Proceedings of 10th Annual Conf. on Genetic and Evolutionary Computation, New York, USA pp. 135–142 (2008)
Teytaud, O., Gelly, S.: General lower bounds for evolutionary algorithms. Proceedings of 9th Conference on Parallel Problem Solving from Nature, pp. 21–31 (2006)
Trelea, I.C.: The particle swarm optimization algorithm: convergence analysis and parameter selection. Inf. Process. Lett. 85, 317–325 (2003)
Vaughan, D.E., Jacobson, S.H., Kaul, H.: Analyzing the performance of simultaneous generalized hill climbing algorithms. Comput. Optim. Appl. 37, 103–119 (2007)
Wegener, I.: Simulated annealing beats metropolis in combinatorial optimization. Proceedings of ICALP ’05, Springer LNCS, Berlin Heidelberg vol. 3580, pp. 589–601 (2005)
Wegener, I., Witt, C.: On the analysis of a simple evolutionary algorithm on quadratic pseudo-boolean functions. J. Discrete Algorithms 3, 61–78 (2005)
Whitley, D., Watson, J.P.: Complexity theory and the no free lunch theorem. In: Burke, E.K., Kendall, G. (eds.) Search Methodologies: Introductory Tutorials in Optimization and Decision Support Techniques, Kluwer, Boston, 317–399 (2005)
Witt, C.: Worst-case and average-case approximations by simple randomized search heuristics. Proceedings of 22nd Annual Symposium on Theoretical Aspects of Computer Science, Springer LNCS, Berlin Heidelberg vol. 3404, pp. 44–56 (2005)
Witt, C.: Runtime analysis of the \((\mu+1)\) EA on simple pseudo-bolan functions. Proceedings of 8th Annual Conference on Genetic and Evolutionary Computation, Seattle, Washington, pp. 651–658 (2006)
Wolpert, D.H., Macready, W.G.: No free lunch theorems for optimization. IEEE Trans. Evol. Comput. 1, 67–82 (1997).
Yao, A.C.: Probabilistic computations: towards a unified measure of complexity. Proceedings of 17th IEEE Symposium on the Foundations of Computer Science, 222–227 (1977)
Yu, Y., Zhou, Z.-H.: A new approach to estimating the expected first hitting time of evolutionary algorithms. Proceedings of 21th National Conference on Artificial Intelligence, Boston, MA, 555–560 (2006)
Zhang, W.: Phase transitions and backbones of the asymmetric travelling salesman problem. J. Artif. Intell. Res. 21, 471–497 (2004)
Zlochin, M., Birattari, M., Meuleau, N., Dorigo, M.: Model-based search for combinatorial optimization: a critical survey. Ann. Oper. Res. 131, 373–379 (2004)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2010 Springer Science+Business Media, LLC
About this chapter
Cite this chapter
Gutjahr, W.J. (2010). Stochastic Search in Metaheuristics. In: Gendreau, M., Potvin, JY. (eds) Handbook of Metaheuristics. International Series in Operations Research & Management Science, vol 146. Springer, Boston, MA. https://doi.org/10.1007/978-1-4419-1665-5_19
Download citation
DOI: https://doi.org/10.1007/978-1-4419-1665-5_19
Published:
Publisher Name: Springer, Boston, MA
Print ISBN: 978-1-4419-1663-1
Online ISBN: 978-1-4419-1665-5
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)