巢湖学院学报 ›› 2018, Vol. 20 ›› Issue (6): 1-7.
• 数学统计 • 下一篇
依据2010 年1 月至2018 年8 月的安徽省实际居民消费价格指数（CPI）数据，利用统计建模的方法，基于非平稳时间序列分析（ARIMA）构建了物价指数预测模型并进行了实证分析。结果显示，该模型的绝对误差及相对绝对误差均位于较小的波动范围之内。说明该模型拟合效果比较理想，预测值逼近真实值。最后，利用该模型对安徽省2018 年9 月至11月的CPI 数据进行了预测。
Based on the actual CPI data of Anhui Province from January 2010 to August 2018, this paper constructs a price index prediction model based on non-stationary time series analysis (ARIMA)and conducts empirical analysis. The results show that the absolute and relative absolute errors of the model are within a small fluctuation range. This shows that the model fits well and the predicted value approaches the true value. Finally, the CPI data from September to November 2018 in Anhui Province is predicted using this model.#br#
