Analiza efektywności wykorzystania kryteriów informacyjnych w prognozowaniu zapotrzebowania na części zamienne

W artykule przedstawiono klasyczne metody prognozowania zapotrzebowania na części zamienne oraz nowy trend w tej dziedzinie - wykorzystanie jednej z metod sztucznej inteligencji - sztucznych sieci neuronowych SSN (Sztuczne Sieci Neuronowe; Artificial Neural Networks, ANN).

The paper presents a new approach to the spare parts forecasting issue - a method which combines regression modeling, information criteria and artificial neural networks ANN. The research presented in this article compares efficiency of classical methods with the artificial intelligence tool in the scope of spare parts forecasting. Artificial Neural Networks have been advocated as an alternative to traditional statistical forecasting methods. Classical methods, such as exponential smoothing or mean average, have been used for several decades in forecasting demand. However, many of these techniques may perform poorly when demand for an item is lumpy or intermittent. In the paper three concepts of using ANN in spare parts forecasting - micro, macro and hybrid - were described. The article presents also the variable selection issue, which is of a great importance in any model building.