SOMA - aplikace

Abstract

The work deals by the searching for optimal solutions using evolutionary algorithms. There is created an application, whose basis consists of Self-Organizing Migrating Algorithm. There were imple-mented five kinds of algorithm at the evolutionary algorithm SOMA– All to One, All to All, All to One Random, All to One Adaptive a All to All Adaptive. All these variants of SOMA algorithm were tested in the gallery of the selected functions. The application offers the possibility to work the more with the best values, which were obtained by different types of SOMA algorithm. Using simulated annealing algorithm, the application tries to find an even better results than were found by SOMA algorithm. In conclusion, the results of various options SOMA algorithm and its experimental exten-sion of simulated annealing are evaluated.