In 1992, Koza proposed genetic programming. It is an expansion to genetic algorithm. Genetic programming characterizes a tree-type non-direct encoding of a probable solution, and computer program might be used in which search is applied directly to the solution. Genetic programming takes up variable-length representation, whereas fixed-length encoding is adopted by genetic algorithm. In genetic programming, population generates variation in values of genes and also in individual structures. Genetic programming involves four steps. They are generation of an initial population for computer programs. It executes each program, depending on how well it solves the problem and assigns fitness value to it. Then, using best currently available programs it creates new computer programs by mutations and crossover.
Leave a Reply