【摘要】 [目的]探讨一种基于秩次的稳健回归分析方法-R估计。[方法]从参数估计及回归诊断方面阐述R估计原理、算法及应用。[结果]误差为正态分布和重尾分布时R估计表现均优于LS估计。[结论]R估计在实际数据偏离假定模型时可以得到客观的估计,有较高的效率,并可抵抗离群点。更多还原

【Abstract】 [Objective]To explore the method of rank-based robust regression-R Estimate.[Methods]The theory of Rank-based regression,its algorithm and the application were discussed.[Results]R estimation showed better behavior than LS estimation under normal distribution and heavy-tailed distribution of errors.[Conclusion]The rank-based methods have excellent efficiency properties over a wide class of error distributions including asymmetric distributions and can decrease the influence of outliers.更多还原