非零截距线性和非线性回归模型的T-型稳健估计算法  

EM Algorithm and Iteration Algorithm for T-Type Estimators in Non-Null Intercept Linear Model and Nonlinear Model

作　　者：王福昌[1] 曹慧荣[2] 钱小仕[1] 

机构地区：[1]中国地震局防灾科技学院,河北三河065201 [2]廊坊师范学院数信学院,河北廊坊065000

出　　处：《数学的实践与认识》2010年第17期169-175,共7页Mathematics in Practice and Theory

基　　金：中国地震局教师科研基金(20090126);河北省教育厅自然科学研究项目(Z2009138)

摘　　要：首先给出非零截距线性模型T-型估计的模型与EM算法,其次给出非线性回归模型参数的T-型估计,利用泰勒级数对模型线性化,得到参数估计的迭代算法,最后用数值模拟实验验证了该算法的正确性和证实了T-型估计的稳健性.At first,the T-type regression estimators of parameters for non-null intercept linear model is discussed,and the EM algorithm for T-type estimators in linear model is given.Subsequently,the T-type regression estimators of parameters for nonlinear model is discussed,and the iteration algorithm by linearizing nonlinear model using Taylor series for T-type estimators in nonlinear model is given.Finally,synthesized and real data examples are conducted to illustrate the validity of algorithms and robustness of T-type estimators.

关 键 词：T-型估计 EM算法 最大似然估计 线性模型 非线性模型 

分 类 号：O212.1[理学—概率论与数理统计]