| With the rapid development of the 3rd party logistics in recent years, the competition is growing increasingly among logistics firms. In the fierce market, a serious problem which firms concern mostly is what customer strategy makes firms more competitive. Decision-making methods like evaluating customers, clustering customers and deciding key customers are able to help logistics firms to constitute customer strategies and draw long-term plans, which eventually improves the strength of the firm in essence. The status of customer can be reflected by business data generated inbusiness activities. Thus the potential of a customer could be acquired via forecasting its business data. As a result, this paper puts forward a various-period forecasting model for seasonal time series which applies appropriate methods to forecast data with various periods. Furthermore, with respect to the data less than 3 years, a seasonal time series forecasting method based on GM(1,1) was advanced to forecast seasonal data precisely. Then a customer clustering model based on K-MEDOIDS was... |