Skip to main content Accessibility help
×
Home

Tight Bounds on the Optimization Time of a Randomized Search Heuristic on Linear Functions

  • CARSTEN WITT (a1)

Abstract

The analysis of randomized search heuristics on classes of functions is fundamental to the understanding of the underlying stochastic process and the development of suitable proof techniques. Recently, remarkable progress has been made in bounding the expected optimization time of a simple evolutionary algorithm, called (1+1) EA, on the class of linear functions. We improve the previously best known bound in this setting from (1.39+o(1))en ln n to en ln n+O(n) in expectation and with high probability, which is tight up to lower-order terms. Moreover, upper and lower bounds for arbitrary mutation probabilities p are derived, which imply expected polynomial optimization time as long as p = O((ln n)/n) and p = Ω(nC) for a constant C > 0, and which are tight if p = c/n for a constant c > 0. As a consequence, the standard mutation probability p = 1/n is optimal for all linear functions, and the (1+1) EA is found to be an optimal mutation-based algorithm. Furthermore, the algorithm turns out to be surprisingly robust since the large neighbourhood explored by the mutation operator does not disrupt the search.

Copyright

Footnotes

Hide All

An extended abstract of this paper appeared at STACS'12 [27].

Footnotes

References

Hide All
[1]Auger, A. and Doerr, B. (2011) Theory of Randomized Search Heuristics: Foundations and Recent Developments, World Scientific.
[2]Bäck, T. (1993) Optimal mutation rates in genetic search. In Proc. International Conference on Genetic Algorithms, ICGA '93, Morgan Kaufmann, pp. 28.
[3]Doerr, B., Fouz, M. and Witt, C. (2010). Quasirandom evolutionary algorithms. In Proc. Genetic and Evolutionary Computation Conference, GECCO '10, ACM Press, pp. 14571464.
[4]Doerr, B., Fouz, M. and Witt, C. (2011). Sharp bounds by probability-generating functions and variable drift. In Proc. Genetic and Evolutionary Computation Conference, GECCO '11, ACM Press, pp. 20832090.
[5]Doerr, B. and Goldberg, L. A. (2010). Adaptive drift analysis. In Proc. Parallel Problem Solving from Nature XI, PPSN '10, Vol. 6238 of Lecture Notes in Computer Science, Springer, pp. 3241.
[6]Doerr, B. and Goldberg, L. A. (2012). Adaptive drift analysis. Algorithmica, in press. Available online: http://www.springerlink.com/content/v0n75x65k016r477
[7]Doerr, B., Johannsen, D. and Winzen, C. (2010). Drift analysis and linear functions revisited. In Proc. Congress on Evolutionary Computation, CEC '10, IEEE Press, pp. 18.
[8]Doerr, B., Johannsen, D. and Winzen, C. (2010). Multiplicative drift analysis. In Proc. Genetic and Evolutionary Computation Conference, GECCO '10, ACM Press, pp. 14491456.
[9]Droste, S., Jansen, T. and Wegener, I. (2000). A natural and simple function which is hard for all evolutionary algorithms. In Proc. 26th Annual Conference of the IEEE Industrial Electronics Society, IECON '00, pp. 27042709.
[10]Droste, S., Jansen, T. and Wegener, I. (2002). On the analysis of the (1+1) evolutionary algorithm. Theoret. Comput. Sci. 276 5181.
[11]Hajek, B. (1982). Hitting-time and occupation-time bounds implied by drift analysis with applications. Adv. Appl. Probab. 13 502525.
[12]He, J. and Yao, X. (2001). Drift analysis and average time complexity of evolutionary algorithms. Artificial Intelligence 127 5785.
[13]He, J. and Yao, X. (2004). A study of drift analysis for estimating computation time of evolutionary algorithms. Natural Computing 3 2135.
[14]Jägersküpper, J. (2007). Algorithmic analysis of a basic evolutionary algorithm for continuous optimization. Theoret. Comput. Sci. 379 329347.
[15]Jägersküpper, J. (2008). A blend of Markov-chain and drift analysis. In Proc. Parallel Problem Solving from Nature X, PPSN '08, Vol. 5199 of Lecture Notes in Computer Science, Springer, pp. 4151.
[16]Jägersküpper, J. (2011). Combining Markov-chain analysis and drift analysis: The (1+1) evolutionary algorithm on linear functions reloaded. Algorithmica 59 409424.
[17]Lehre, P. K. and Witt, C. (2012). Black-box search by unbiased variation. Algorithmica 64 623642.
[18]Marshall, A. W., Olkin, I. and Arnold, B. (2011). Inequalities: Theory of Majorization and its Applications, second edition, Springer.
[19]Motwani, R. and Raghavan, P. (1995). Randomized Algorithms, Cambridge University Press.
[20]Mühlenbein, H. (1992). How genetic algorithms really work I: Mutation and hillclimbing. In Proc. Parallel Problem Solving from Nature I, PPSN '92, Elsevier, pp. 1526.
[21]Neumann, F. and Witt, C. (2010). Bioinspired Computation in Combinatorial Optimization: Algorithms and Their Computational Complexity, Natural Computing Series, Springer.
[22]Sudholt, D. (2010). General lower bounds for the running time of evolutionary algorithms. In Proc. Parallel Problem Solving from Nature XI, PPSN '10, Vol. 6238 of Lecture Notes in Computer Science, Springer, pp. 124133.
[23]Sudholt, D. (2012). A new method for lower bounds on the running time of evolutionary algorithms. IEEE Trans. Evolutionary Computation, to appear. Preprint: http://dx.doi.org/10.1109/TEVC.2012.2202241
[24]Wegener, I. (2001). Methods for the analysis of evolutionary algorithms on pseudo-Boolean functions. In Evolutionary Optimization (Sarker, R., Mohammadian, M. and Yao, X., eds), Kluwer.
[25]Wegener, I. and Witt, C. (2005). On the analysis of a simple evolutionary algorithm on quadratic pseudo-boolean functions. J. Discrete Algorithms 3 6178.
[26]Wegener, I. and Witt, C. (2005). On the optimization of monotone polynomials by simple randomized search heuristics. Combin. Probab. Comput. 14 225247.
[27]Witt, C. (2012). Optimizing linear functions with randomized search heuristics: The robustness of mutation. In Proc. Symposium on Theoretical Aspects of Computer Science, STACS '12, Leibniz International Proceedings in Informatics, pp. 420431. Open access: http://drops.dagstuhl.de/opus/volltexte/2012/3392/.

Keywords

Related content

Powered by UNSILO

Tight Bounds on the Optimization Time of a Randomized Search Heuristic on Linear Functions

  • CARSTEN WITT (a1)

Metrics

Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed.