回归预测模型的稳健性讨论——以润扬大桥工程为例

Robust Estimation of Regression Calculation Model——By the Example of the Runyang Bridge Engineering

经典的LS估计法(least squares estimation,LS estimation)是对每个观测数据都给予相同的权重,由此带来了对于异常值的处理不当,从而影响了回归模型的有效性,本文采用迭代加权最小二乘法来进行稳健估计(RobustEstimation),并基于MATLAB软件结合润扬大桥工程实例说明了此法的有效性。

Classical least squares estimation gives the same weight to every observation which results in inappropriate handling of anomalous values and affecting feasibility of regression calculation model. This paper uses iterative weighted least square method based on MATLAB soft for makingrobust estimation by the example of the Runyang Bridge engineering.