摘要
通过一个案例——中国农村居民人均消耗函数,介绍了回归诊断和最小一乘准则在建立计量经济学模型中的应用,在数据中有异常点或强影响点存在的情况下,应采用稳健性较优的最小一乘准则给出参数的估计.
The application of the regression diagnosis and least-absolute criteria in the establishment of econometric models were introduced.The least absolute criteria with better robustness should be adopted for parameter estimation when there are outliers or influence cases in the data.
出处
《内蒙古科技大学学报》
CAS
2007年第3期280-283,共4页
Journal of Inner Mongolia University of Science and Technology
关键词
异常点
强影响点
最小一乘准则
outlier
influence case
least-absolute crieria