摘要
本文研究上海市近期粮食价格变化情况。通过近期36个月的上海市和全国粮食消费价格指数进行对比和统计分析,发现曲线没有明显趋势且平缓,两者高度相关。利用真实数据对上海市近期96周的粮食价格(以重点监测的批发市场优质粳米成交均价为例)深入分析,并进行白噪声检验。原序列经过一阶差分和5期的移动平均平滑处理之后,继续模型识别,建立ARIMA(2,1,2)模型,并用ARIMA(2,1,2)预测未来10期的价格。模型优化后解决残差序列中存在的异方差问题,最终得到完整的拟合模型。
This paper studies the recent changes of grain prices in Shanghai. By comparing and analyzing the recent 36 months’ grain consumer price index between Shanghai and the whole country, it is found that there is no obvious trend, and both curves are gentle and highly correlated. The real data were used to analyze the grain price of Shanghai in recent 96 weeks (taking the average transaction price of high-quality Rice in the wholesale market as an example) in Shanghai, and applied the white noise test to this time series. After one order differencing and 5-period moving average smoothing of the original series, the ARIMA (2,1,2) model is established and the price of the next 10 periods is predicted. After the optimization of the model, the heteroscedasticity problem in the residual sequence is solved, and finally a complete fitting model is obtained.
出处
《统计学与应用》
2020年第4期684-695,共12页
Statistical and Application
