摘要
本文讨论了含有一个自变量且自变量误差方差与因变量误差方差不等的变量合误差线性回归模型。由于对这种模型最大似然估计失去作用，文中提出了一种新的估计方法──驻点似然估计。这种方法就是，用使似然函数达到驻点的参数值作为参数真值的估计。作为此估计法的实际应用，本文导出了所讨论模型的参数的驻点似然估计。然后证明了所得到的估计在一定条件下是模型参数的强相合估计。
In this paper,a error in variables linear regression model of two variables,where Error variances of the two variables are unequal, is discussed.A new estimating method,stationary point estimation,is sugested.The method uses the values of parameters at which the likelihood function reaches its stationary point,as the estimator of the true values of the model parameters.As its practice,the stationary point likelihood estimator of the present model is derived.And the strong consistency of this estimation of parameters is proved under certain conditions.
出处
《工程数学学报》
CSCD
1996年第1期83-88,共6页
Chinese Journal of Engineering Mathematics
