< Back to Article

Virtual Patients and Sensitivity Analysis of the Guyton Model of Blood Pressure Regulation: Towards Individualized Models of Whole-Body Physiology

Figure 8

Evaluation of linear classifiers for identifying hypertensive virtual individuals.

Each classifier (binomial GLM) was fitted to a random sample of the virtual population and then evaluated on the remaining . A: ROC curves for several classifiers; the optimal (30-parameter) classifier has an area under curve (AUC) of , demonstrating high predictive power. The 6-parameter “Renal+Liver” classifier performs nearly as well (AUC = 0.948). B: The parameter sensitivity of the optimal classifier. The y-axis measures the variation in the prediction over the range of values for each parameter.

doi: https://doi.org/10.1371/journal.pcbi.1002571.g008