Abstract
It is known that evolution strategies in continuous domains might not converge in the presence of noise [3, 14]. It is also known that, under mild assumptions, and using an increasing number of resamplings, one can mitigate the effect of additive noise [4] and recover convergence. We show new sufficient conditions for the convergence of an evolutionary algorithm with constant number of resamplings; in particular, we get fast rates (log-linear convergence) provided that the variance decreases around the optimum slightly faster than in the so-called multiplicative noise model.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsPreview
Unable to display preview. Download preview PDF.
References
Arnold, D., Beyer, H.-G.: Investigation of the \((\mu ,\lambda )\)-es in the presence of noise. In: Proc. of the IEEE Conference on Evolutionary Computation (CEC 2001), pp. 332–339. IEEE (2001)
Arnold, D., Beyer, H.-G.: Local performance of the (1 + 1)-es in a noisy environment. IEEE Transactions on Evolutionary Computation 6(1), 30–41 (2002)
Arnold, D.V., Beyer, H.-G.: A general noise model and its effects on evolution strategy performance. IEEE Transactions on Evolutionary Computation 10(4), 380–391 (2006)
Astete-Morales, S., Liu, J., Teytaud, O.: log-log convergence for noisy optimization. In: Proceedings of EA 2013. LNCS. Springer (2013) (page accepted)
Auger, A.: Convergence results for (1,\(\lambda \))-SA-ES using the theory of \(\varphi \)-irreducible Markov chains. Theoretical Computer Science 334(1–3), 35–69 (2005)
Auger, A.: Linear convergence on positively homogeneous functions of a comparison-based step-size adaptive randomized search: the (1+1)-es with generalized one-fifth success rule (2013) (submitted)
Auger, A., Jebalia, M., Teytaud, O.: (x, sigma, eta): quasi-random mutations for evolution strategies. In: EA, p. 12 (2005)
Beyer, H.-G.: The Theory of Evolution Strategies. Natural Computing Series. Springer, Heidelberg (2001)
Chen, H.: Lower rate of convergence for locating the maximum of a function. Annals of statistics 16, 1330–1334 (1988)
Coulom, R.: CLOP: Confident Local Optimization for Noisy Black-Box Parameter Tuning. In: van den Herik, H.J., Plaat, A. (eds.) ACG 2011. LNCS, vol. 7168, pp. 146–157. Springer, Heidelberg (2012)
Fabian, V.: Stochastic Approximation of Minima with Improved Asymptotic Speed. Annals of Mathematical statistics 38, 191–200 (1967)
Fabian, V.: Stochastic Approximation. SLP. Department of Statistics and Probability, Michigan State University (1971)
Heidrich-Meisner, V., Igel, C.: Hoeffding and bernstein races for selecting policies in evolutionary direct policy search. In: ICML 2009: Proceedings of the 26th Annual International Conference on Machine Learning, pp. 401–408. ACM, New York (2009)
Jebalia, M., Auger, A., Hansen, N.: Log linear convergence and divergence of the scale-invariant (1+1)-ES in noisy environments. Algorithmica (2010)
Rechenberg, I.: Evolutionstrategie: Optimierung Technischer Systeme nach Prinzipien des Biologischen Evolution. Fromman-Holzboog Verlag, Stuttgart (1973)
Shamir, O.: On the complexity of bandit and derivative-free stochastic convex optimization. CoRR, abs/1209.2388 (2012)
Teytaud, O., Decock, J.: Noisy Optimization Complexity. In: FOGA - Foundations of Genetic Algorithms XII - 2013, Adelaide, Australie (February 2013)
Teytaud, O., Fournier, H.: Lower Bounds for Evolution Strategies Using VC-Dimension. In: Rudolph, G., Jansen, T., Lucas, S., Poloni, C., Beume, N. (eds.) PPSN 2008. LNCS, vol. 5199, pp. 102–111. Springer, Heidelberg (2008)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Cauwet, ML. (2014). Noisy Optimization: Convergence with a Fixed Number of Resamplings. In: Esparcia-Alcázar, A., Mora, A. (eds) Applications of Evolutionary Computation. EvoApplications 2014. Lecture Notes in Computer Science(), vol 8602. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-662-45523-4_49
Download citation
DOI: https://doi.org/10.1007/978-3-662-45523-4_49
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-662-45522-7
Online ISBN: 978-3-662-45523-4
eBook Packages: Computer ScienceComputer Science (R0)