Evolution of evolutionary algorithms

Objašnjena je ideja evolucije evolucijskih algoritama linearnim genetskim programiranjem. Evolucijski algoritam predstavljen je kao niz instrukcija na dva načina. Razvijen je algoritam genetskog programiranja koji evoluira evolucijski algoritam s obzirom na to koliko dobro evolucijski algoritam rješava problem za koji je napravljen. Navedni algoritam ostvaren je u programskom jeziku Java. Program je ispitan na primjeru evolucijskog algoritma za optimizaciju funkcija te rezultati pokazuju da je ovako razvijen evolucijski algoritam usporediv, ponekad i bolji, od standardnog ručno napisanog evolucijskog algoritma.The idea of evolving evolutionary algorithms using linear genetic programming is explained. Evolutionary algorithm is represented as a list of instructions in two ways. Genetic programming algorithm, which evolves an evolutionary algorithm taking into consideration how well the evolutionary algorithm solves the problem it was developed for, is developed. It is implemented in Java programming language. The program is tested using the evolutionary algorithm for function optimization and results show that evolutionary algorithm developed in this way has results similar, or sometimes better, than the standard manually written evolutionary algorithm