约束下线性混合模型参数估计的渐近解及其收敛性  

Approximate Solution and Convergence of Parameter Estimation of a Line Mixed model with Line Inequality Constraints

作　　者：段智力[1] 

机构地区：[1]白城师范学院数学系,吉林白城137000

出　　处：《数学的实践与认识》2012年第4期186-194,共9页Mathematics in Practice and Theory

基　　金：国家自然科学基金(10701021)

摘　　要：在线性不等式约束下讨论了具有相同参数的两个线性混合模型的参数估计问题,给出了一种迭代算法,得到了这类模型中参数的最小二乘估计序列及其渐近解.在此基础上,利用多元多项式方程组解的个数定理和不动点定理,证明了此估计序列是依概率1收敛的.This paper provides algorithm for parameters estimation question of math model, which contains two line mixed models with the same parameters with linear inequality con- straints, furthermore, gaining two estimation sequences that refer to the parameters in the model ,On the basis of it, Using fixed-point theorem and numbers of solutions multivariate polynomial equation group proves estimation sequences which are convergent in probability 1.

关 键 词：约束线性混合模型 最小二乘估计序列 收敛性 

分 类 号：O212.1[理学—概率论与数理统计]