Aiming at data characteristics of noise existing in the product sale series, a support vector machine based on the adaptive segmented loss function(AS??-SVM) is proposed. In the AS??-SVM, a separate insensitive loss value is assigned to each sample point adaptively, which can reduce the influence of inaccurate samples on the final model. It is proved in theory that the method can enhance partial generalization performance of the model. The AS??-SVM is applied to a numerical value example and the automobile sales forecasting in contrast with the ?? support vector machine (??-SVM). The experiment results
show that the AS??-SVM is effective and feasible, by which more accurate forecasting results are obtained over the ??-SVM.