[1]成立花,张俊敏.新的递推有界GM回归估计算法[J].华侨大学学报(自然科学版),2015,36(3):359-364.[doi:10.11830/ISSN.1000-5013.2015.03.0359]
　CHENG Li-hua,ZHANG Jun-min.A New Recursive Bounded GM Estimator for Regression[J].Journal of Huaqiao University(Natural Science),2015,36(3):359-364.[doi:10.11830/ISSN.1000-5013.2015.03.0359]
点击复制

新的递推有界GM回归估计算法()
分享到：

《华侨大学学报（自然科学版）》[ISSN:1000-5013/CN:35-1079/N]

卷:
第36卷
期数:
2015年第3期
页码:
359-364
栏目:
出版日期:
2015-05-20

文章信息/Info

Title:
A New Recursive Bounded GM Estimator for Regression
文章编号:
1000-5013(2015)03-0359-06
作者:
成立花1;  张俊敏2
1. 西安工程大学 理学院, 陕西 西安 710048;2. 西安建筑科技大学 理学院, 陕西 西安 710055
Author(s):
CHENG Li-hua1;  ZHANG Jun-min2
1. College of Science, Xi’an Polytechnic University, Xi’an 710048, China; 2. College of Science, Xi’an University of Architecture and Technology, Xi’an 710055, China
关键词:
GM估计器;  鲁棒估计;  AR模型;  加性异常点
Keywords:
generalized maximum likelihood type stimator;  robust estimation;  autoregressive parameters;  additive outliers
分类号:
TN911.7;O241.2;N945.14
DOI:
10.11830/ISSN.1000-5013.2015.03.0359
文献标志码:
A
摘要:
提出一种新的递推有界广义极大似然类(GM)回归估计器,新估计器所用的风险函数基于更一般的框架,并采用有界M-估计函数.设计一个新的权函数拒绝或降低异常点对估计结果的影响,并增加一个增广项,提出一种具有较强自适应能力的面向自回归(AR)模型参数估计的算法.仿真结果表明:提出的GM回归估计器及面向AR模型的算法对异常点不利影响(主要来自于回归变量中的加性异常点)的抑制效果均优于其他GM估计器;在参数不做任何调整的情况下,面向AR模型的算法对非平稳环境下的估计具有良好的估计精度和收敛性.
Abstract:
A new recursive bounded GM estimator for regression is proposed. Unlike other GM estimators, the new estimator is based on one more general framework and uses a cost function with bounded M-estimate function. The new estimator,in effect, is a recursive one-step iteration solution of the "normal equations" corresponding to the cost function. In the new estimator, a weight function is designed to reject or to reduce the influence of the outliers. Furthermore, by introducing an augment variable, the proposed estimator is modified to a very adaptive version for the estimation of autoregressive parameters. The simulation results show that both the proposed estimator and its modification are more effective than other related estimators in suppressing the adverse influence of outliers; the proposed estimator, with the same settings, can keep a high accuracy and stable convergence performance in a variety of non-stationary environments.

参考文献/References:

[1] HUBER P J.Robust regression: Asymptotics, conjectures and monte carlo[J].Annals of Statistics, 1973, 1(5):799-821.
[2] CAMPBEL K.Recursive computation of M-estimates for the parameters of a finite autoregressive process[J].The Annals of Stat,1982,10(2):442-453.
[3] ANTOCH J,EKBLOM H.Recursive robust regression computational aspects and comparison[J].Computational Statistics and Data Analysis,1995,19(2):115-128.
[4] SEJLING K,et al.Methods for recursive robust estimation of AR parameters[J].Computational Atatistics and Data Analysis,1994,17(5):509-536.
[5] PHAM D S,ZOUBIR A M.A sequential algorithm for robust parameter estimation[J].IEEE Signal Processing Lett,2005,12(1):21-24.
[6] VEGA L R,REY H,BENESTY J,et al.A robust recursive least squares algorithm[J].IEEE Trans Signal Process,2009,57(3):1209-1216.
[7] KRASKER W S,WELSCH R E.Efficient bounded-influence regression estimation[J].Journal of the American Statistical Association,1982,77(379):595-604.
[8] GRILLENZONI C.Recursive generalized M-estimators of system parameters[J].Technometrics,1997,39(2):211-224.
[9] ENGIUND J E.Recursive versions of the algorithm by Krasker and Welsch[J].Sequential Analysis,1991,10(3/4):211-234.
[10] MARONNA R A,MARTIN R D,YOHAI V J.Robust statistics: Theory and methods[M].West Sussex:John Wiley&Sons,2006:888-889.
[11] ROUSSEEUW P J,LEROY A M.Robust regression and outlier detection[M].New York:Wiley,1987:340-347.

备注/Memo

备注/Memo:
收稿日期: 2015-01-11
通信作者: 成立花(1973-),女,副教授,主要从事领域泛函分析、估计理论及应用的研究.E-mail:178529238@qq.com.
基金项目: 国家自然科学基金资助项目(11101323); 陕西省教育厅自然科学专项基金(12JK0879)
更新日期/Last Update: 2015-05-20