Figure 2.
Performance of medNet under different simulation and model settings. This study assesses medNet's functionality across logistic regression (LR), support vector machine (SVM), and random forest (RF) models, in three distinct simulation scenarios. Consistent T (0.6) and H (18) settings were maintained, and each experiment was repeated 20 times. (A) The averaged true positive rate (TPR) and false positive rate (FPR) values for exposures, along with the scaled averages of TPR, FPR, and false discovery rate (FDR) for mediators, categorized by model_sample size_scenario and (B) the distributions of TPR, FPR, and FDR values for mediators.
