缺失数据下线性模型中反映变量均值的经验似然置信区间

Empirical Likelihood Confidence Interval for the Mean of Response Variables in Linear Regression Models with Missing Data

在协变量和反映变量都缺失下,构造了线性模型中反映变量均值的经验似然置信区间,数据模拟表明调整的经验似然置信区间有较好的覆盖率和精度,进一步完善了缺失数据下对线性模型的研究.

The empirical likelihood confidence interval is constructed for the mean of response variables in linear regression model when the independent and the response variables all have missing. Our simulation indicates that the adjusted empirical likelihood confidence regions have good coverage accuracies and average lengths, which perfected the study of linear regression model with missing data.