Fig. 6From: Development of a predictive model to distinguish prostate cancer from benign prostatic hyperplasia by integrating serum glycoproteomics and clinical variablesROC curve analysis of all the 4 tested ML algorithms (Random forest, blue curve; Logistic Regression, orange curve; KNN classifier, green curve; Decision Tree classifier, red curve) and the confusion matrix relative to the best performing ML algorithm (Random Forest)Back to article page