摘要
以1992-2012年的安徽省城镇居民人均消费支出的数据,运用多元回归与时间序列结合的模型预测、ARIMA模型预测和灰色预测三种单项预测方法,以预测的误差平方和最小为准则,建立IOWA组合预测模型,并以安徽省城镇人均消费支出为例进行实证分析,发现组合预测模型在整体上都优于每一单项预测方法,对我国居民消费支出预测和研究城镇居民人均消费具有重大意义。
Based on the analysis of urban residents per capita consumption expenditure data from 1992 to 2012 in Anhui province and by using the combination of multiple regression and time series model,ARIMA model prediction and gray prediction three single forecasting method and the minimum prediction error sum of squares as the criterion,IOWA combination forecast model is set up.The author of this paper makes a case study of urban per capita consumption expenditure in Anhui province.The author finds that the combination forecast model on the whole is better than that of each single prediction methods and holds that to improve the residents' consumption expenditure and the research of urban residents' per capita consumption is of great significance.
出处
《怀化学院学报》
2014年第3期32-35,共4页
Journal of Huaihua University
基金
国家社科基金项目"组合预测模型与方法创新及其优化理论研究"(12BTJ008)
关键词
人均消费支出
ARIMA模型预测
灰色预测
组合预测
the per capita consumption expenditure
ARIMA model forecast
Grey prediction
combination forecast