摘要
ARIMA模型作为统计预测中的一种重要模型,被广泛运用于各个领域中。本文在非季节性及季节性ARIMA模型的基础上介绍了乘积ARIMA模型的一般表达式,提供了建立这类模型及使用其进行预报的一般过程,并以某超市月顾客量数据进行了实例分析。分析结果表明,当一时间序列不仅含有季节性成分,还混有非季节性成分时,用乘积ARIMA模型的预测精度较高。但该模型只适用于短期预测。
ARIMA model is widely used in many fields as a model for statistic prediction. Base on seasonal and non-seasonal ARIMA models, this paper presents a general formula of multiply ARIMA model, providing a general progress to establish and apply the model. Moreover, an example of customers in one supermarket was applied. The date show that multiply ARIMA model has a higher forecast precision when time series including seasonal and non-seasonal components simultaneously. However, this model only adopts to short period prediction.
出处
《华北科技学院学报》
2008年第2期85-89,共5页
Journal of North China Institute of Science and Technology
