A demand forecasting method integrating qualitative and quantitative approaches and its application in cigarette inventory

Reasonable demand forecasting is vital for an enterprise to determine adequate production to deal with market requirements and maintain low-level inventory. We propose an integrated demand forecasting model that combines qualitative and quantitative methods to forecast specific demands. Within the framework of the model, we propose a qualitative forecasting method that combines analytic hierarchical processes and evidence theory, and a seasonal quadratic exponential smoothing (SQES) method that corrects predicted results by multiplying a seasonal factor. The proposed model is applied to the demand for cigarette inventory of a tobacco company in China. Several quantitative methods are implemented to produce forecast results, while the qualitative method is performed to obtain a forecast interval. The proposed model is demonstrated to be capable of forecasting the actual demand reliably.