Immune Genetic Programming and it's application in bioprocess modeling

The canonical form of Genetic Programming (GP) relies almost exclusively on the crossover operator for exploring the solution space. But the crossover operator makes individuals change randomly and indirectly during the whole process, it not only give the individuals the evolutionary chance but also cause certain degeneracy. Here, based on the analysis of the cross-operating characteristics and the study of individual structure, a new form of vaccine and a new autonomous obtaining method is proposed. The concept of immunity is integrated into the GP and a novel algorithm - immune GP is proposed. Experiments show that the result obtained by the proposed algorithm restrains the degenerative phenomena during evolution, so as to improve the algorithm efficiency. © 2013 TCCT, CAA.