ITERATIVE QUASI-LIKELIHOOD FOR SEEMINGLY UNRELATED REGRESSION SYSTEMS  

作　　者：LINLU FANYUNZHENG DUJUN YUANYUAN 

机构地区：[1]SchoolofMathematicsandSystemSciences,ShandongUniversity,Jinan250100,Shandong,China. [2]SchoolofMathematicsandSystemSciences,ShandongUniversity,Jinan250100,Shandong,China／／DepartmentofMathematics,NantongUniversity,Nantong226007,Jiangsu,China. [3]SchoolofMathematicalSciences,NankaiUniversity,Tianjin300071,China. [4]ProjectsupportedbytheNationalNaturalScienceFoundationofChina

出　　处：《Chinese Annals of Mathematics,Series B》2005年第3期335-346,共12页数学年刊（B辑英文版）

基　　金：Project supported by the National Natural Science Foundation of China (No.10371059, No.10171051).

摘　　要：In the seemingly unrelated regression systems, the existing quasi-likelihood is always involved in the difficult problem of calculating inverse of a high order matrix specially for large systems. To avoid this problem, the new quasi-likelihood proposed in this paper is based mainly on a linearly iterative process of some unbiased estimating functions.Some finite sample properties and asymptotic behaviours for this new quasi-likelihood are investigated. These results show that the new quasi-likelihood for parameter of interest is E-sufficient, iteratively efficient and approximately efficient. Some examples are given to illustrate the theoretical results.

关 键 词：Seemingly unrelated regression system QUASI-LIKELIHOOD Unbiased estimating function 

分 类 号：O212.1[理学—概率论与数理统计]