Description: Comparative analysis of mathematical models used for prediction of Complex system’s state is formed. According to Results of work the using of several kind models has an advan tages before using of along model. The main result of this urticle is a functional decomposition of IT for adaptive modelling. IT includes two kind of models: analytical model based on Regression equation and Volters differential equations; artificial neural-network. The main problems of IT for adaptive modelling is formed.
Keywords: mathematical model, complex system, artificial neural-network, diversification