Skip to main content
. 2019 Nov 15;98(46):e17510. doi: 10.1097/MD.0000000000017510

Figure 4.

Figure 4

Comparison between RF and LR predictive models performed on sub-dataset 1. RF predictive model performed better than LR model on sub-dataset 1. All metrics represent the average over the 10-folds of the cross validation. AuROC = area under the receiver operating characteristic curve, LR = logic regression, RF = random forest, TN rate = true negative rate.