摘要
讨论二种类型的污染回归 .一是 (Ⅰ )型污染回归模型 ,并不假定误差分布为正态 ,也不假定误差方差为已知 ,只假定误差有三阶矩 ,在此条件下证明了污染系数及回归系数的参数估计的强相合性 ;二是提出一种新的污染回归模型 (Ⅲ ) ,在只假定误差有二阶矩 (并不假定有已知方差 )的条件下 ,得到污染系数及回归系数的参数估计的相合性 .
This paper studies two kinds of contamination regression models.The first is model (Ⅰ).We neither set error distribution being normal distribution nor set variance of error being known.We only set error having third moment and set up strong consistence of contamination coefficient and regression coefficient.The second is a new contamination regression model (Ⅲ).Suppose that the error has only the second moment (but not to assume it′s variance being known),consistence of estimations of contamination coefficient and regression coefficients can be obtained.
出处
《同济大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2001年第6期703-707,共5页
Journal of Tongji University:Natural Science
关键词
污染系数
污染数据
相合估计
可识别条件
半参数回归模型
污染回归模型
参数估计
contamination coefficients
contamination data
strong consistent estimate
consistent estimate
recognizable condition
semi parameter regression model
