Forecasting the Efficiency of Innovative Industrial Systems Based on Neural Networks

Approaches presented today in the scientific literature suggest that there are no methodological solutions based on training of artificial neural networks to predict direction industrial development, taking into account a set factors—innovation, environmental friendliness, modernization and production growth. The aim study is develop predictive model performance management innovative systems by building networks. research methods were correlation analysis, (species—regression), extrapolation, exponential smoothing. As result research, estimation efficiency technique an system complex considering criteria technical modernization, activity, ecologization developed; prognostic network models allow optimize contribution signs formation target (set) values indicators for macro micro-industrial will level growth trajectory systems; priority directions their development offered. following conclusions: determined volume sales goods, products waste recycling, which allows save resources; results forecasting depend significantly DataSet formulated. Although multilayer independently select important features, it advisable conduct analysis beforehand, provide higher probability high-quality model. novelty lies testing unique methodology assess effectiveness systems: multidimensional approach (takes factors innovation, growth); combines number tools (correlation, ranking weighting); expands method assessment terms composition variables (previously approaches limited aspects considered).