OATAO - Open Archive Toulouse Archive Ouverte Open Access Week

Surrogate modeling approximation using a mixture of experts based on EM joint estimation

Bettebghor, Dimitri and Bartoli, Nathalie and Grihon, Stéphane and Morlier, Joseph and Samuelides, Manuel Surrogate modeling approximation using a mixture of experts based on EM joint estimation. (2010) Structural and Multidisciplinary Optimization . ISSN 1615-1488

[img] (Document in English)

PDF (Publisher's version) - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
1MB

Official URL: http://dx.doi.org/10.1007/s00158-010-0554-2

Abstract

An automatic method to combine several local surrogate models is presented. This method is intended to build accurate and smooth approximation of discontinuous functions that are to be used in structural optimization problems. It strongly relies on the Expectation-Maximization (EM) algorithm for Gaussian mixture models (GMM). To the end of regression, the inputs are clustered together with their output values by means of parameter estimation of the joint distribution. A local expert is then built (linear, quadratic, artificial neural network, moving least squares) on each cluster. Lastly, the local experts are combined using the Gaussian mixture model parameters found by the EM algorithm to obtain a global model. This method is tested over both mathematical test cases and an engineering optimization problem from aeronautics and is found to improve the accuracy of the approximation.

Item Type:Article
Additional Information:Thanks to Springer editor. The definitive version is available at http://www.springer.com
Audience (journal):International peer-reviewed journal
Uncontrolled Keywords:
Institution:Other partners > Airbus (FRANCE)
Université de Toulouse > Institut Supérieur de l'Aéronautique et de l'Espace - ISAE
French research institutions > Office National d'Etudes et Recherches Aérospatiales - ONERA
Laboratory name:
Statistics:download
Deposited By: Joseph Morlier
Deposited On:18 Oct 2010 13:45

Repository Staff Only: item control page