缺失数据下均值与方差模型的统计诊断

Statistical Diagnosis for Mean and Variance Models with Missing Data

摘要: 针对响应变量随机缺失下均值与方差模型，考虑了基于数据删除模型的统计诊断方法。其中主要基于回归插补法和随机回归插补法以及结合Gauss-Newton迭代计算算法给出该模型中未知参数的极大似然估计，进而基于似然距离进行异常值诊断分析。最后通过模拟研究分析，结果表明所提出的模型和统计方法是可行有效的。

Abstract: A statistical diagnosis method based on the data deletion model is considered for the mean and variance models with response variables random missing. It is mainly based on the regression imputation and random regression imputation and the Gauss-Newton iterative algorithm to give the maximum likelihood estimation of the unknown parameters in the models, and then based on the likelihood distance to carry out the diagnosis and analysis of the abnormal values. Finally, through simulation analysis, the results show that the proposed model and statistical method are feasible and effective.