Evolutionary identification and synthesis of predictive models

Abstract

In this contribution is demonstrated use of two evolutionary algorithms on parameter identification of selected predictive models. Both algorithms were used to indentify parameter of pre-selected ARMA models. At the end are discussed possibilities of use of synthesis of predictive models by means of methods of symbolic regression that has successfully been used on chaotic system identification by means of evolutionary algorithms on measured data. © 2013 Springer-Verlag.