简单线性回归模型的分离参数估计方法

Calculation of separate parameter estimation of simple linear regression model

最小二乘法的参数估计方法虽然体现了拟合结果的整体优化特征，却没有体现出各参数估计量的自身统计特征，对此本文给出一种新的参数估计方法，同最小二乘法相比，该方法得到的参数估计值受统计数据的影响最小，文中还讨论了参数估计值自身具有的线性和无偏性。

The value of parameter estimation in the least square method can express the whole majorizing character of fitting solutionn,but cannot express the individual statistic character of itself.A new calculation method is proposed,which can make the estimators be affected by statistic data within the least extent com-pared with the least square method.The linearity and unbiasedness of parame-ter estimators is also discussed.