Fig. 7From: Federated learning algorithms for generalized mixed-effects model (GLMM) on horizontally partitioned data from distributed sourcesThe ROC curve with Area Under Curve (AUC) among centralized, Laplace, and Gauss–Hermite methods. The orange ROC curve is the centralized method without regularization and the Laplace approximation(i.e., R implementation in the ‘lme4’ package, which does not have an option for including regularization). AUC values are also included, a higher AUC value implicates better performance of the model. The green ROC curve is the 2-degree Gauss–Hermite method with regularizationBack to article page