Global optimization, differential evolution, self-adaptation of control parameters, nonlinear regression.

Adaptation of control parameters in differential evolution is considered. Five adaptive variants of differential evolution are applied to the estimation of parameters in nonlinear regression models and their performance is compared experimentally with the adaptive controlled random search algorithm tailored especially for these problems. The NIST nonlinear regression datasets are used as a benchmark. Two of five tested variants of adaptive differential evolution perform almost as reliable as the adaptive controlled random search algorithm and one of these two variants converges only slightly slower and its time requirements are almost comparable with the adaptive controlled random search.

Adaptation of control parameters in differential evolution is considered. Five adaptive variants of differential evolution are applied to the estimation of parameters in nonlinear regression models and their performance is compared experimentally with the adaptive controlled random search algorithm tailored especially for these problems. The NIST nonlinear regression datasets are used as a benchmark. Two of five tested variants of adaptive differential evolution perform almost as reliable as the adaptive controlled random search algorithm and one of these two variants converges only slightly slower and its time requirements are almost comparable with the adaptive controlled random search.