"5"^^ . . . "Model of multilayer artificial neural network for prediction of iron ore demand" . "978-80-87294-24-6" . . "Besta, Petr" . "With regards to many effects which can disrupt the delivery schedule of iron ore to a company, this raw material must be ordered well in advance. Determination of the required order volume results from prediction of iron ore demand. With respect to the fluctuations on the metallurgical commodity market, it is very difficult to use classical prediction models based on time series analysis. A typical example when models based solely on historical time series can not be used to predict iron ore demand is the world economic crisis period, because the demand for metallurgical commodities witnesses a sharp decrease. The article presents prediction model based on multilayer artificial neural network which takes into account not only historical data of iron ore demand but also information regarding the current situation on the world steel market and the iron ore stock volume of a given metallurgical company. The model is designed in such a way that the output is represented by the demand prediction of iron ore for the next month. Levenberg-Maguardt algorithm was used for learning the network. The proposed model should be included in the hybrid intelligent decision support system which will make it possible to efficiently reduce uncertainty and risk of logistics decision-making in the sphere of iron ore supply." . . "S, Z(MSM6198910015)" . "27360" . . . "Model of multilayer artificial neural network for prediction of iron ore demand"@en . "212897" . . . . "With regards to many effects which can disrupt the delivery schedule of iron ore to a company, this raw material must be ordered well in advance. Determination of the required order volume results from prediction of iron ore demand. With respect to the fluctuations on the metallurgical commodity market, it is very difficult to use classical prediction models based on time series analysis. A typical example when models based solely on historical time series can not be used to predict iron ore demand is the world economic crisis period, because the demand for metallurgical commodities witnesses a sharp decrease. The article presents prediction model based on multilayer artificial neural network which takes into account not only historical data of iron ore demand but also information regarding the current situation on the world steel market and the iron ore stock volume of a given metallurgical company. The model is designed in such a way that the output is represented by the demand prediction of iron ore for the next month. Levenberg-Maguardt algorithm was used for learning the network. The proposed model should be included in the hybrid intelligent decision support system which will make it possible to efficiently reduce uncertainty and risk of logistics decision-making in the sphere of iron ore supply."@en . "Model of multilayer artificial neural network for prediction of iron ore demand" . "2011-05-18+02:00"^^ . . "Model of multilayer artificial neural network for prediction of iron ore demand"@en . "Ostrava" . . "Tanger s.r.o." . "Brno" . "3"^^ . . . "20th Anniversary International Conference on Metallurgy and Materials: METAL 2011" . . "Feliks, J." . . "Lenort, Radim" . "artificial neural network; iron ore demand; prediction models"@en . "RIV/61989100:27360/11:86081883!RIV12-MSM-27360___" . "[5F1EA6ED6F47]" . "2"^^ . . "RIV/61989100:27360/11:86081883" . . . .