Creating evolutionary algorithms by means of analytic programming-Preliminary study

Abstract

This contribution deals with a new idea how to create evolutionary algorithms by means of symbolic regression and Analytic Programming. The motivation was not only to tune some existing algorithms to their better performance, but also to find a new robust evolutionary algorithm. In this preliminary study operators of Differential Evolution-Select DE, MutaeDE and CroseDE were used during a process of Analytic Programming. The results showed that AP was able to find the originally defined DE but also new structure which has a similar behaviour but slower convergence in multimodal function than DE. This, in further work, leads to including the conditions of convergence to CostFunction. Results were showed in 100 repeated simulations in graphical and also table way.