BOOSTERS: a derivative-free algorithm based on radial basis functions

• Published in: International Journal of Modelling and Simulation, vol. 29, num. 1, p. 26-36
• Publication date: 2009

Derivative-free optimization involves the methods used to minimize an expensive objective functionwhen its derivatives are not available. We present here a trust-region algorithmbased on Radial Basis Functions (RBFs). The main originality of our approach is the use of RBFs to build the trust-region models and our management of the interpolation points based on Newton fundamental polynomials. Moreover the complexity of ourmethod is very attractive. We have tested the algorithmagainst the best state-of-theart methods (UOBYQA, NEWUOA, DFO). The tests on the problems from the CUTEr collection show that BOOSTERS is performing very well on medium-size problems. Moreover, it is able to solve problems of dimension 200, which is considered very large in derivative-free optimization.

Reference

• TRANSP-OR-ARTICLE-2008-013
• doi:10.2316/Journal.205.2009.1.205-4634

Record created on 2006-02-14, modified on 2012-03-21