Application of an combination forecasting modal in logistics demand

Analyses the influence factors of logistics demand, make use of regression analyses, grey forecasting and neural network to construct single forecasting model. For some limitations existing in the single models, this paper put forward the combination-forecasting model for logistics demand based on the least error absolute value. Verify the modal with concrete example. The result of error analyses shows that it considerably improves precision of forecast and manifests the validities for forecasting of logistics demand.