江苏省商品零售价格指数时间序列分析
时间:2020-06-21 14:31 来源:毕业论文 作者:毕业论文 点击:次
| 摘要 本文使用江苏省1979-2014 年的商品的零售价格指数作为时间序列数据.首 先对数据进行预处理，即对数据的平稳性和纯随机性进行判断，得出该序列是一个平稳非白噪声序列的结论；然后对该序列数据建立 ARMA 模型，在模型建立过程中发现MA模型和AR模型均显著有效，所以两个模型都适用；由于模型不唯一，所以用AIC 准则和SBC准则来优化模型， 选择使AIC值和SBC 值均最小的模型为相对最优模型；最后选择 MA 模型来对江苏省未来三年的商品零售价格指数进行预测. 该论文有图15 幅，表8 个，参考文献5篇. 51392 毕业论文关键词：商品零售价格指数 ARMA 模型 AIC 准则 SBC 准则 时间序列分析 Time Series Analysis of Retail Price Index in Jiangsu Abstract This paper is based on retail price index from 1979 to 2014 in Jiangsu Province.First of all,is the data preprocessing, which is stable and pure random data to determine that the sequence is a non stationary white noise series. Then through this data to establish the ARMA model. In the model building process found in the MA model and AR model were significantly effective, so the two models can be used.Because the model is not unique, so use AIC and SBC criteria to select the optimal model.Choosing the model that make the AIC and SBC are all smallest as the relative optimum model. Finally, I choose the MA model to predict the next three years of retail price index in Jiangsu Province. Key Words: Retail price index ARMA model AIC criterion SBC criterion Time series analysis 目录 摘要Ⅲ Abstract-Ⅳ 目录V 图清单-VI 表清单-VI 1绪论1 2ARMA模型1 2.1AR模型-1 2.2MA模型2 2.3ARMA模型2 3建模步骤-2 3.1序列数据预处理3 3.2模型定阶-3 3.3参数估计-3 3.4模型检验-4 3.5模型优化-5 4对商品零售价格指数建模-5 4.1序列数据预处理6 4.2模型定阶-8 4.3参数估计-9 4.4模型检验10 4.5模型优化12 4.6模型预测12 5小结与结论-13 参考文献-14
致谢-15 |