摘要:
本文把随机效应看作缺失数据并利用P-样条拟合非参数部分,应用MCNR算法得到了半参数非线性再生散度混合效应模型的未知参数的估计,同时利用Q函数得到了模型的广义Cook距离。此外,本文还研究了三种不同扰动情形的局部影响分析,得到了相应的影响矩阵。最后,通过模拟和实例验证了本文所提出的估计方法的有效性。
Abstract: This paper proposes several case-deletion and local influence measures for assessing the influence of an observation for semiparametric nonlinear reproductive dispersion mixed-effects models. The essential idea is to treat the latent random effects in the model as missing data and estimate unknown parameters by MCNR algorithm. On the basis of the Q-function which is associated with the conditional expectation of the complete-data log-likelihood, we generate generalized Cook Distance. Three different perturbation schemes are discussed. Finally, the proposed methods are illustrated by the simulation analysis and one real data set.