Forecasting demand of short life cycle products by SVM

In this paper, the demand of short life cycle products is forecasted by the method of SVM in the conditions of data deficiency. The method considers productspsila demand, demand forecasted by Bass model and season factor as affecting factors of the demand of short life cycle product, the training sample and forecasting sample varies when the time changes. Then SVM forecasting model is set up and with it the demand is forecasted. The comparison with relative models indicates that the presented method is more valid in forecasting the demand of short life cycle products.