#
Efficient global optimization: testing, reliability and efficiency

###
Invited seminar at Université Paul Sabatier: 14th October 2003

###
J. A. J. Hall, K. I. M. McKinnon and T. Mayer

####
Abstract

In many practical optimization problems, the number of function
evaluations is severely limited by time or cost. This practical
consideration has driven the development of efficient methods for global
optimization which require only small numbers of function evaluations.
This talk will consider, in particular, the method of Jones {\em et al.}
in which the objective is modelled by a linear predictor which
interpolates the function at a set of sample points. A corresponding
standard error function models the uncertainty in the predictor at
points not yet sampled. By optimizing a merit function of the predictor
and standard error, the best new sample point is determined. A
procedure for generating appropriate test problems for such methods will
be described. Extensive testing of the method has allowed its
reliability to be properly assessed and techniques for its improvement
will be discussed. The optimization of the merit function is, itself, a
global optimization problem and the scope for its efficient solution
will be examined.

**Slides:**

Postscript EGO_14.10.03.ps

Compressed Postscript EGO_14.10.03.ps.Z

G-Zipped Postscript EGO_14.10.03.ps.gz

PDF EGO_14.10.03.pdf
Note that the postscript version is of finer resolution.

Last modified: