论文部分内容阅读
本文使用蒙特卡罗方法,求得广义线性混合模型之最大似然估计,并提供用来评估统计参数之收敛和精确度之实用方法．仿真研究显示无偏之固定效应参数估计,而方差分量估计之误差则相近于前人结果．应用举例为使用泊松分布求取乳癌死亡率之小区域估计．
In this paper, we use Monte Carlo method to obtain the maximum likelihood estimation of generalized linear mixed model and provide a practical method to evaluate the convergence and accuracy of statistical parameters. Simulation studies show unbiased fixed-effects parameter estimates, while errors in variance component estimates are similar to those of previous generations. Application examples for the use of Poisson distribution of small breast cancer mortality estimates.