A New Estimator for Regression Model with Serially Correlated Errors
The problem of estimating the unknown parameter in the econometric model with serial correlation is considered. It was found that the estimator βof βis near its true value,whereas the estimator ρ of ρis relatively far from its true value. A new estimator is given under the sum of squared residuals and the absolute residuals criteria respectively .The method is illustrated by a simulation study.It is pointed out that the key idea of the model is to improve the estimator of ρ .
Junxing Zhou
Mathematics and Statistics School Zhejiang University of Finance and Economics Hangzhou Zhejiang China 310018
国际会议
南宁
英文
2007-07-20(万方平台首次上网日期,不代表论文的发表时间)