Estimation for a class of linear regression models
GAN Shengjin1,2，WANG Qiongjin3
1.School of Electronical and Information Engineering， Fuqing Branch of Fujian Normal University， Fuqing，Fujian 350300，China； 2.Department of Statistics， Chengchi University，Taipei 11605，China；3.Department of Statistics of Faculty， Economy and Management， East China Normal University， Shanghai 200062， China
Abstract：In this paper， a class of linear models with multi-populations are considered. They have a partial regressional coefficients in common. Weight average of all populations's ordinary least squares is proposed to estimate the same partial regressional coefficients， which is then put into each linear regressional models to estimate the other coefficients by ordinary least squares. It is verified that the estimators proposed are not only unbiased， but also have smaller convariance matraces compared with estimators by each populations's ordinary least squares. Simulations demonstrated that our method is efficient.
甘胜进,王琼瑾. 一类线性回归模型的参数估计[J]. 华中师范大学学报(自然科学版), 2021, 55(3): 351-355.
GAN Shengjin,WANG Qiongjin. Estimation for a class of linear regression models. journal1, 2021, 55(3): 351-355.
2021, Vol. 55 
