摘要
在ARIMA模型和BP神经网络两种单一预测模型的基础上,根据组合预测误差平方和最小原则构建一种线性组合预测模型,并将其应用于销售预测中。实证结果表明,该组合预测模型能有效提高销售预测精度,具有良好的应用前景。
Using ARIMA time series model and BP neural network, this paper has set up the combined model of prediction, and discussed the key problem, how to determine the weights of different methods, which could forecast the future need of products. Furthermore, armed on defining the volume of sales of the corporation in the future, based on the numbers demanded in the past, we made a prediction of a corporation using the combined forecasting model. The research shows that this combined model, could improve the precision of prediction and minish the error obviously, and has a good application prospects.
基金
国家自然科学基金资助项目(70671017)
