Advertisement

An informational approach to the global optimization of expensive-to-evaluate functions

  • Julien Villemonteix
  • Emmanuel Vazquez
  • Eric Walter
Article

Abstract

In many global optimization problems motivated by engineering applications, the number of function evaluations is severely limited by time or cost. To ensure that each evaluation contributes to the localization of good candidates for the role of global minimizer, a sequential choice of evaluation points is usually carried out. In particular, when Kriging is used to interpolate past evaluations, the uncertainty associated with the lack of information on the function can be expressed and used to compute a number of criteria accounting for the interest of an additional evaluation at any given point. This paper introduces minimizers entropy as a new Kriging-based criterion for the sequential choice of points at which the function should be evaluated. Based on stepwise uncertainty reduction, it accounts for the informational gain on the minimizer expected from a new evaluation. The criterion is approximated using conditional simulations of the Gaussian process model behind Kriging, and then inserted into an algorithm similar in spirit to the Efficient Global Optimization (EGO) algorithm. An empirical comparison is carried out between our criterion and expected improvement, one of the reference criteria in the literature. Experimental results indicate major evaluation savings over EGO. Finally, the method, which we call IAGO (for Informational Approach to Global Optimization), is extended to robust optimization problems, where both the factors to be tuned and the function evaluations are corrupted by noise.

Keywords

Gaussian process Global optimization Kriging Robust optimization Stepwise uncertainty reduction 

References

  1. 1.
    Abrahamsen, P.: A review of Gaussian random fields and correlation functions. Tech. Rep., Norwegian Computing Center (1997). www.math.ntnu.no/~omre/TMA4250/V2007/abrahamsen2.ps
  2. 2.
    Adler R.: On excursion sets, tubes formulas and maxima of random fields. Ann. Appl. Prob. 10(1), 1–74 (2000)Google Scholar
  3. 3.
    Chib S., Greenberg E.: Understanding the metropolis-hastings algorithm. Am. Stat. 49(4), 327–335 (1995)CrossRefGoogle Scholar
  4. 4.
    Chilès J., Delfiner P.: Geostatistics, Modeling Spatial Uncertainty. Wiley, New York (1999)Google Scholar
  5. 5.
    Cover T.M., Thomas A.J.: Elements of Information Theory. Wiley, New York (1991)CrossRefGoogle Scholar
  6. 6.
    Cox, D., John, S.: Sdo: a statistical method for global optimization. In: Alexandrov, N., Hussaini, M.Y. (eds.) Multidisciplinary Design Optimization: State of the Art, pp. 315–329. SIAM, Philadelphia (1997). citeseer.ifi.unizh.ch/cox97sdo.html
  7. 7.
    Delfiner, P.: Shift invariance under linear models. Ph.D. thesis, Princetown University, New Jersey (1977)Google Scholar
  8. 8.
    Dixon, L., Szegö, G.: The global optimisation problem: an introduction. In: Dixon, L., Szegö, G. (eds.) Towards Global Optimization 2. North-Holland Publishing Company (1978)Google Scholar
  9. 9.
    Geman, D., Jedynak, B.: An active testing model for tracking roads in satellite images. Tech. Rep. 2757, Institut National de Recherche en Informatique et en Automatique (INRIA) (1995)Google Scholar
  10. 10.
    Gutmann H.: A radial basis function method for global optimization. J. Glob. Optim. 19(3), 201–227 (2001)CrossRefGoogle Scholar
  11. 11.
    Huang, D.: Experimental planning and sequential Kriging optimization using variable fidelity data. Ph.D. thesis, Ohio State University (2005)Google Scholar
  12. 12.
    Huang D., Allen T., Notz W., Zeng N.: Global optimization of stochastic black-box systems via sequential Kriging meta-models. J. Glob. Optim. 34, 441–466 (2006)CrossRefGoogle Scholar
  13. 13.
    Jones D.: A taxonomy of global optimization methods based on response surfaces. J.Glob. Optim. 21, 345–383 (2001)CrossRefGoogle Scholar
  14. 14.
    Jones D., Schonlau M., William J.: Efficient global optimization of expensive black-box functions. J. Glob. Optim. 13, 455–492 (1998)CrossRefGoogle Scholar
  15. 15.
    Matheron G.: Principles of geostatistics. Econ. Geol. 58, 1246–1266 (1963)CrossRefGoogle Scholar
  16. 16.
    Matheron, G.: Le krigeage universel. In: Cahiers du Centre de Morphologie Mathématique de Fontainebleau. Ecole des Mines de Paris (1969) Fasc.1Google Scholar
  17. 17.
    Sasena M., Papalambros P., Goovaerts P.: Exploration of metamodeling sampling criteria for constrained global optimization. Eng. Opt. 34, 263–278 (2002)CrossRefGoogle Scholar
  18. 18.
    Schonlau, M.: Computer experiments and global optimization. Ph.D. thesis, University of Waterloo (1997)Google Scholar
  19. 19.
    Sjö, E.: Crossings and maxima in Gaussian fields and seas. Ph.D. thesis, Lund Institute of Technology (2000)Google Scholar
  20. 20.
    Smola, A.: Learning with kernels. Ph.D. thesis, Technische Universität Berlin (1998)Google Scholar
  21. 21.
    Stein M.: Interpolation of Spatial Data: Some Theory for Kriging. Springer, New York (1999)Google Scholar
  22. 22.
    Vechia A.: Estimation and model identification for continuous spatial processes. J. R. Stat. Soc. B(50), 297–312 (1998)Google Scholar
  23. 23.
    Wahba G.: Support vector machines, reproducing kernel Hilbert spaces, and randomized GACV. In: Schölkopf, B., Burges, C., Smola, A. (eds) Advances in Kernel Methods—Support Vector Learning, pp. 69–87. MIT Press, Boston (1998)Google Scholar
  24. 24.
    Watson A., Barnes R.: Infill sampling criteria to locate extremes. Math. Geol. 27(5), 589–698 (1995)CrossRefGoogle Scholar
  25. 25.
    Williams, C., Rasmussen, C.: Gaussian processes for regression. In: Touretzky, D., Mayer, M., Hasselmo, M. (eds.) Advances in Neural Information Processing Systems, vol. 8. MIT Press (1996)Google Scholar
  26. 26.
    Williams B., Santner T., Notz W.: Sequential design of computer experiments to minimize integrated response functions. Stat. Sinica 10, 1133–1152 (2000)Google Scholar
  27. 27.
    Yaglom A.: Correlation Theory of Stationary and Related Random Functions I: Basic Results. Vol. 6, Springer Series in Statistics. Springer-Verlag, New-York (1986)Google Scholar

Copyright information

© Springer Science+Business Media, LLC. 2008

Authors and Affiliations

  • Julien Villemonteix
    • 1
  • Emmanuel Vazquez
    • 2
  • Eric Walter
    • 3
  1. 1.Energy Systems DepartmentRenault S.A.GuyancourtFrance
  2. 2.SUPELECGif-sur-YvetteFrance
  3. 3.Laboratoire des Signaux et SystèmesCNRS-SUPELEC-Univ Paris-SudGif-sur-YvetteFrance

Personalised recommendations