A Trust-Region Algorithm for Global Optimization

Bernardetta Addis (b.addising.unifi.it)
Sven Leyffer (leyffermcs.anl.gov)

Abstract: We consider the global minimization of a bound-constrained function with a so-called funnel structure. We develop a two-phase procedure that uses sampling, local optimization, and Gaussian smoothing to construct a smooth model of the underlying funnel. The procedure is embedded in a trust-region framework that avoids the pitfalls of a fixed sampling radius. We present a numerical comparison to three popular methods and show that the new algorithm is robust and uses up to 20 times fewer local minimizations steps.

Keywords: Global optimization, smoothing, trust region

Category 1: Global Optimization (Stochastic Approaches )

Citation: Mathematics and Computer Science Division Preprint ANL/MCS-P1190-0804, Argonne National Laboratory, Argonne, IL, August 2004

Download: [Postscript][PDF]

Entry Submitted: 08/06/2004
Entry Accepted: 08/06/2004
Entry Last Modified: 08/13/2004

Modify/Update this entry