关联时间序列异常数据的预警模型研究

Caution model about exceptional data of the correlative time series

在贝叶斯时间序列动态模型的基础上,建立起一种关联时间序列上异常数据的预警模型,并在商品价格监控领域进行了量价实证分析研究.结果表明,该模型能有效解决此类关联时间序列异常数据的隐蔽性和时效性问题.

A caution model about exceptional data of the correlative time series based on the Bayesian forecasting and dynamic models was set up.The model was applied to commodity price surveillance,and the results showed that the concealment and timeliness of such exceptional data of the correlative time series could be solved effectively by the model.