时间序列分析中的ARMA算法及其软件实现

ARMA Method of the Time Series Analysis and the Implement With the Software

阐述了一种适用于经济领域中的统计预测方法──ＡＲＭＡ法。它将预测对象随时间变化的序列，看作是一个随机序列。运用相应的数学模型加以近似描述，利用自相关分析这一最有效的工具，通过对建立模型、识别模型、检验模型、预测模型的深入研究，更本质地认识了这些动态数据的内在结构，并运用ＲＡＴＳ软件对该模型成功地编制了一个界面友好、交互性强的实用汉化程序，达到了最佳预测效果，对于经济统计预测有一定的实用价值。

This article focuses on one of statistics forecast methods in the field of economics──ARMA. It regards the series with the change of time as the radam series. It uses the suitable math model and the most efficient tool autocorrelation analysis to describe. The structure of data can be realized essentially by studying the establishment, identification, testing, forecasting of the model and accomplishment of Chinese application of the programme with the RATS. This application achieves the best forecasting effect and has the practical value in the statistics forecast of the field of the economics .