Recombination guidance for numerical genetic programming

In the authors´ earlier papers (1993, 1994, 1995), they introduced an adaptive program called STROGANOFF (STructured Representation On Genetic Algorithms for NOnlinear Function Fitting), which integrated a multiple regression analysis method and a genetic algorithm-based search strategy. The effectiveness of STROGANOFF was demonstrated by solving several system identification problems. This paper proposes an adaptive recombination mechanism for STROGANOFF. The intention is to exploit already-built structures by adaptive recombination, in which genetic program recombination is guided by a certain measure. The effectiveness of this approach is shown by an experiment in predicting a chaotic time series. The authors describe real-world applications of STROGANOFF to computer vision