传统的最小二乘估计在处理含有异常值的线性回归模型的估计问题时拟合得往往不好,根据GL算法给出稳定加权最小二乘优化的参数的估计法.并结合实例进行回归分析,验证该方法的高精度和高效率.
It is often dissatisfactory to deal with the estimation of the parameters of the general linear regression model with traditional Ieast square estimate when some abnormal data need to be analyzed. By using the GL Algorithm on weighed parameters, a new stable center is obtained to linear regression model. Finally, an applied example is given, which verifies the precision and effectiveness of this method.