A Hybrid Optimization Method for Neural Tree Network Model

摘要

Neural tree network model has been successfully applied to solving large numbers of complex nonlinear problems in control area. The optimization of neural tree model contains: structure and parameter, the major problem in evolving structure without parameter information was noisy fitness evaluation problem, so an improved genetic programming algorithm is proposed to synthesize the optimization process. Simulation results on two time series prediction problems show that the proposed strategy is a potential method with better performance and effectiveness.