混合效应模型在葡萄糖耐糖量测试中的应用研究

Application of Mixed Effect Model in Glucose Tolerance Test

摘要: 本文是对一组来自标准葡萄糖耐糖量实验的平衡纵向数据构建线性混合效应模型、lasso回归和广义线性混合模型，利用AIC准则以及lasso的变量选择选取最优的模型，并对所得模型进行模拟预测，然后计算模型的均方误差并进行比较，选出均方误差相对较小的模型。发现线性混合模型在这组数据中具有较好的预测效果。本文所有计算均用R软件完成。

Abstract: This article is about a group of glucose sugar levels in experimental equilibrium longitudinal resistance from standard data to construct linear mixed effects models, lasso regression and generalized linear mixed models, using the AIC criterion as well as the lasso variable selection to select the optimal model and forecast the model simulated, then the mean square error of the model is calculated and compared, and the model with relatively small mean square error is selected. It is found that the linear mixed model has a good predictive effect in this set of data. All calculations in this paper are completed by R software.