R. Oeuvray and M. Bierlaire (Switzerland)
Optimization, derivative-free optimization, radial basis
functions, medical image registration problem.
Derivative-free optimization involves all the methods used
to minimize an objective function when its derivatives are
not available and when the function is expensive. We
present here a trust region algorithm based on radial basis
functions instead of the classical second order polynomi
als. Actually, our surrogate of the objective function is a
mixed radial and polynomial model. We present results on
the medical image registration problem. On the instances
of this problem, our method surpasses all the tested state
of-the-art derivative-free algorithms.