摘要
主要考虑了同方差型的半参数线性回归模型中参数的随机加权最小二乘估计(RWLSE).讨论了用随机加权Bootstrap方法来逼近LSE的分布,证明这种逼近是以概率1渐近有效的.
The randomly weighted least square estimator （RWLSE） for the parametric component in semiparametric regression models was mainly discussed. The randomly weighted bootstrap method was used to approximate the distribution of the least square estimators （LSE）, and it was demonstrated that the approximation is asymptoticly valid with probability one.
基金
国家自然科学基金(10471136)
中国科学院知识创新重要方向项目(KJCX3-SYW-S02)资助
关键词
半参数回归模型
渐近正态性
随机加权最小二乘估计
semi-parametric regression models
asymptotic normality properties
randomly weighted least square estimator
