Robust Gaussian Process-Based Global Optimization Using a Fully Bayesian Expected Improvement Criterion
We consider the problem of optimizing a real-valued continuous function f, which is supposed to be expensive to evaluate and, consequently, can only be evaluated a limited number of times. This article focuses on the Bayesian approach to this problem, which consists in combining evaluation results and prior information about f in order to efficiently select new evaluation points, as long as the budget for evaluations is not exhausted.
The algorithm called efficient global optimization (EGO), proposed by Jones, Schonlau and Welch (J. Global Optim., 13(4):455–492, 1998), is one of the most popular Bayesian optimization algorithms. It is based on a sampling criterion called the expected improvement (EI), which assumes a Gaussian process prior about f. In the EGO algorithm, the parameters of the covariance of the Gaussian process are estimated from the evaluation results by maximum likelihood, and these parameters are then plugged in the EI sampling criterion. However, it is well-known that this plug-in strategy can lead to very disappointing results when the evaluation results do not carry enough information about f to estimate the parameters in a satisfactory manner.
We advocate a fully Bayesian approach to this problem, and derive an analytical expression for the EI criterion in the case of Student predictive distributions. Numerical experiments show that the fully Bayesian approach makes EI-based optimization more robust while maintaining an average loss similar to that of the EGO algorithm.
KeywordsGlobal Optimization Covariance Function Bayesian Approach Gaussian Process Evaluation Point
Unable to display preview. Download preview PDF.
- 6.Mockus, J., Tiesis, V., Zilinskas, A.: The application of Bayesian methods for seeking the extremum. In: Dixon, L., Szego, G. (eds.) Towards Global Optimization, vol. 2, pp. 117–129. Elsevier, Amsterdam (1978)Google Scholar
- 12.Grünewälder, S., Audibert, J.-Y., Opper, M., Shawe-Taylor, J.: Regret bounds for Gaussian process bandit problems. In: Proceedings of the 13th International Conference on Artificial Intelligence and Statistics (AISTATS 2010). JMLR W&CP, vol. 9, pp. 273–280 (2010)Google Scholar
- 15.Forrester, A.I.J., Jones, D.R.: Global optimization of deceptive functions with sparse sampling. In: 12th AIAA/ISSMO Multidisciplinary Analysis and Optimization Conference, September 10-12 (2008)Google Scholar
- 17.Osborne, M.A.: Bayesian Gaussian Processes for Sequential Prediction Optimisation and Quadrature. PhD thesis, University of Oxford (2010)Google Scholar
- 18.Osborne, M.A., Garnett, R., Roberts, S.J.: Gaussian processes for global optimization. In: 3rd International Conference on Learning and Intelligent Optimization (LION3), Online Proceedings, Trento, Italy (2009)Google Scholar
- 19.Osborne, M.A., Roberts, S.J., Rogers, A., Ramchurn, S.D., Jennings, N.R.: Towards real-time information processing of sensor network data using computationally efficient multi-output Gaussian processes. In: Proceedings of the 7th International Conference on Information Processing in Sensor Networks, pp. 109–120. IEEE Computer Society, Los Alamitos (2008)Google Scholar
- 21.Schonlau, M.: Computer experiments and global optimization. PhD thesis, University of Waterloo, Waterloo, Ontario, Canada (1997)Google Scholar
- 22.Schonlau, M., Welch, W.J.: Global optimization with nonparametric function fitting. In: Proceedings of the ASA, Section on Physical and Engineering Sciences, pp. 183–186. Amer. Statist. Assoc. (1996)Google Scholar
- 23.Schonlau, M., Welch, W.J., Jones, D.R.: A data analytic approach to Bayesian global optimization. In: Proceedings of the ASA, Section on Physical and Engineering Sciences, pp. 186–191. Amer. Statist. Assoc. (1997)Google Scholar
- 33.O’Hagan, A.: Some Bayesian numerical analysis. In: Bayesian Statistics 4: Proceedings of the Fourth Valencia International Meeting, April 15-20, 1991. Oxford University Press, Oxford (1992)Google Scholar