Product Service Demand Forecasting in Hierarchical Service Structure

The problem of product service forecasting in hierarchical service structure is common in many contexts. It is common that the results of the higher level forecasting are in conflict with the sum of the lower level forecasts. The phenomenon has confused the managers in decision making. This aim of this research is to forecast the product services demands in a hierarchical service structure context. By integrating the methods combination forecasting approach and the information combination forecasting approach, the paper provide a multi-level combined forecasting model with six steps. A case of air compressor services forecasting is employed to verify the developed approach. The in-sample and out-of-sample test indicates that the proposed model outperforms the individual models and the direct combination of individual methods. The proposed method could be easily customized for solving other product service forecasting problems, especially when the hierarchical time series data are involved.