Abstract
The aim of this work is to compare different approaches for parallelization in model-based optimization. As another alternative aside from the existing methods, we propose using a multi-objective infill criterion that rewards both the diversity and the expected improvement of the proposed points. This criterion can be applied more universally than the existing ones because it has less requirements. Internally, an evolutionary algorithm is used to optimize this criterion. We verify the usefulness of the approach on a large set of established benchmark problems for black-box optimization. The experiments indicate that the new method’s performance is competitive with other batch techniques and single-step EGO.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
References
Koch, P., Bischl, B., Flasch, O., Bartz-Beielstein, T., Weihs, C., Konen, W.: Tuning and evolution of support vector kernels. Evol. Intel. 5(3), 153–170 (2012)
Hutter, F., Hoos, H.H., Leyton-Brown, K.: Sequential model-based optimization for general algorithm configuration. In: Coello, C.A.C. (ed.) LION 5. LNCS, vol. 6683, pp. 507–523. Springer, Heidelberg (2011)
Hess, S., Wagner, T., Bischl, B.: PROGRESS: progressive reinforcement-learning-based surrogate selection. In: Nicosia, G., Pardalos, P. (eds.) LION 7. LNCS, vol. 7997, pp. 110–124. Springer, Heidelberg (2013)
Jones, D.R., Schonlau, M., Welch, W.J.: Efficient global optimization of expensive black-box functions. J. Global Optim. 13(4), 455–492 (1998)
Krige, D.: A statistical approach to some basic mine valuation problems on the witwatersrand. J. Chem. Metall. Min. Soc. S. Afr. 52(6), 119–139 (1951)
Locatelli, M.: Bayesian algorithms for one-dimensional global optimization. J. Global Optim. 10(1), 57–76 (1997)
Jones, D.R.: A taxonomy of global optimization methods based on response surfaces. J. Global Optim. 21(4), 345–383 (2001)
Vazquez, E., Bect, J.: Convergence properties of the expected improvement algorithm with fixed mean and covariance functions. J. Stat. Plann. Infer. 140(11), 3088–3095 (2010)
Ginsbourger, D., Le Riche, R., Carraro, L.: Kriging is well-suited to parallelize optimization. In: Tenne, Y., Goh, C.-K. (eds.) Computational Intel. in Expensive Opti. Prob. ALO, vol. 2, pp. 131–162. Springer, Heidelberg (2010)
Janusevskis, J., Le Riche, R., Ginsbourger, D., Girdziusas, R.: Expected improvements for the asynchronous parallel global optimization of expensive functions: potentials and challenges. In: Hamadi, Y., Schoenauer, M. (eds.) LION 6. LNCS, vol. 7219, pp. 413–418. Springer, Heidelberg (2012)
Chevalier, C., Ginsbourger, D.: Fast computation of the multi-points expected improvement with applications in batch selection. In: Nicosia, G., Pardalos, P. (eds.) LION 7. LNCS, vol. 7997, pp. 59–69. Springer, Heidelberg (2013)
Hutter, F., Hoos, H.H., Leyton-Brown, K.: Parallel algorithm configuration. In: Hamadi, Y., Schoenauer, M. (eds.) LION 6. LNCS, vol. 7219, pp. 55–70. Springer, Heidelberg (2012)
Coello, C.A.C., Lamont, G.B., Van Veldhuizen, D.A.: Evolutionary Algorithms for Solving Multi-Objective Problems, 2nd edn. Springer, New York (2007)
Brockhoff, D., Friedrich, T., Hebbinghaus, N., Klein, C., Neumann, F., Zitzler, E.: Do additional objectives make a problem harder? In: Proceedings of the 9th Annual Conference on Genetic and Evolutionary Computation, GECCO ’07, pp. 765–772. ACM (2007)
Tran, T.D., Brockhoff, D., Derbel, B.: Multiobjectivization with NSGA-II on the Noiseless BBOB Testbed. In: GECCO (Companion), Workshop on Black-Box Optimization Benchmarking (BBOB’2013), Amsterdam, Pays-Bas, July 2013
Ulrich, T., Thiele, L.: Maximizing population diversity in single-objective optimization. In: Proceedings of the 13th Annual Conference on Genetic and Evolutionary Computation, GECCO ’11, pp. 641–648. ACM, New York (2011)
Wessing, S., Preuss, M., Rudolph, G.: Niching by multiobjectivization with neighbor information: Trade-offs and benefits. In: IEEE Congress on Evolutionary Computation (CEC), pp. 103–110 (2013)
Beume, N., Naujoks, B., Emmerich, M.: SMS-EMOA: Multiobjective selection based on dominated hypervolume. Eur. J. Oper. Res. 181(3), 1653–1669 (2007)
Hansen, N., Finck, S., Ros, R., Auger, A.: Real-parameter black-box optimization benchmarking 2009: noiseless functions definitions. Technical report RR-6829, INRIA (2009). http://hal.inria.fr/inria-00362633/en
Mersmann, O., Bischl, B., Bossek, J., Judt, L.: soobench: Single Objective Optimization Benchmark Functions. R package version 1.1-164
Bischl, B.: mlr: Machine Learning in R. R package version 1.2
Conover, W.: Practical Nonparametric Statistics, 2nd edn. Wiley, New York (1980)
Acknowledgements
This paper is based on investigations of the projects B3 and C2 of the Collaborative Research Center SFB 823, which are kindly supported by Deutsche Forschungsgemeinschaft (DFG). It is also partly supported by the French national research agency (ANR) within the Modeles Numeriques project NumBBO. The authors also thank Tobias Wagner for fruitful discussions of multiobjective infill criteria.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2014 Springer International Publishing Switzerland
About this paper
Cite this paper
Bischl, B., Wessing, S., Bauer, N., Friedrichs, K., Weihs, C. (2014). MOI-MBO: Multiobjective Infill for Parallel Model-Based Optimization. In: Pardalos, P., Resende, M., Vogiatzis, C., Walteros, J. (eds) Learning and Intelligent Optimization. LION 2014. Lecture Notes in Computer Science(), vol 8426. Springer, Cham. https://doi.org/10.1007/978-3-319-09584-4_17
Download citation
DOI: https://doi.org/10.1007/978-3-319-09584-4_17
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-09583-7
Online ISBN: 978-3-319-09584-4
eBook Packages: Computer ScienceComputer Science (R0)