基于乘积ARIMA模型的产品不确定性需求预测

为对不确定的市场需求进行有效预测,可运用SAS系统中的TimeSeriesForecastingSystem,基于乘积求和自回归平均滑动模型(ARIMA),对产品销售的时间序列数据进行预测。先对原始数据或变换后的数据作简单差分或季节差分,把时间序列化为平稳的时间序列,进行参数的初估计,然后进行多次拟合并最终确定模型,根据SAS估计结果,可得到预测偏差。对预测结果在营销管理和供应链管理中的应用进行了分析,取得了较好的效果。

Prediction of the uncertain demand for product by a multiple ARIMA model

It's a critical problem for enterprises, management to predict the uncertain demand for product, and it's very important for optimization of management. Especially, it's a key process for an alliance of supply chain to optimizing the performance by predicting the uncertain demand and sharing the information. A multiple ARIMA model is used to solve this problem by time series forecasting system in SAS. The prediction result and its applications in the marketing management and supply chain management are discussed.