The paper deals with the process of modification the simulation model forecasting development processes of regional macroeconomic systems by constructing an artificial neuron network − a multilayer perceptron with the paradigm of "learning with a teacher". For this type of network is formed the vector of input data, is selected the activation function and the learning algorithm. The multilayer perceptron will solve the problem of determining the parameters that are used to form the parameters and variables of the model in predicting the effects of state policy.
macroeconomic system, forecasting, simulation model, neuron network, state policy
"Prohnozyrovanye protsessov razvytyia rehyonalnыkh makroэkonomycheskykh system s pomoshchiu neironnoi sety" ,
Information Processing Systems,