Forecasting Supply Chain Demand Approach Using Knowledge Management Processes and Supervised Learning Techniques

In today's context (competition and knowledge economy), ML and KM on the supply chain level have received increased attention aiming to determine long and short-term success of many companies. However, demand forecasting with maximum accuracy is absolutely critical to invest in various fields, which places the knowledge extract process in high demand. In this paper, the authors propose a hybrid approach of prediction into a demand forecasting process in supply chain based on the one hand, on the processes analysis for best professional knowledge for required competencies. And on the other hand, the use of different data sources by supervised learning to improve the process of acquiring explicit knowledge, maximizing the efficiency of the demand forecasting, and comparing the obtained efficiency results. Therefore, the results reveal that the practices of KM should be considered as the most important factors affecting the demand forecasting process in supply chain. The classifier performance is examined by using a confusion matrix based on their accuracy and Kappa value.