Aerospace Instrument-Making Annotation << Back Neural network approximation of gas-turbine engine model based on entropy approach V.I. Vasiljev, S.S. Valeev The problem of gas-turbine engine model approximation based on entropy approach is considered. The results of comparative analysis of spline and neural network interpolation show that in the latter case the entropy of model is less. The example of neural network structure design for multi-mode gas-turbine engine based on compromise between a complexity of neural network and its accuracy is presented.