摘要
对于国内生产总值的预测问题,运用贝叶斯估计方法将主观先验信息、样本拟合区信息和单个模型的预测信息进行了有效综合,导出了三种组合预测方法,并给出了误差方差的估计量.最后是一个应用实例.
In order to predict domestic total output value, we present three combinatorial forecasting algorithms by use of Bayesian method in which subjective prior information and information selected from sample-fitted-district or single forecasting model are effectively synthesized, then give corresponding estimation of error variance following each algorithm. At last a practical example is made to illustrate how to use these three combinatorial algorithms.
出处
《数学的实践与认识》
CSCD
北大核心
2007年第7期1-6,共6页
Mathematics in Practice and Theory
基金
河南省自然科学基金项目(994053200)
关键词
国内生产总值
组合预测
贝叶斯方法
估计量
domestic total output volue
combinatorial forecasting
Bayesian method