Skip to main content

Table 6 Performance of the three ML algorithms using the 5 variables dataset with bootstrapping and of the HellenicSCORE, against the 10-year CVD incidence of ATTICA study participants

From: Machine learning methodologies versus cardiovascular risk scores, in predicting disease risk

Algorithm Accuracy Specificity Sensitivity PPV NPV
k-NN 0.66 0.62 0.67 0.91 0.26
RF 0.79 0.47 0.85 0.90 0.37
DT 0.78 0.48 0.84 0.90 0.36
HellenicSCORE 0.85 0.20 0.97 0.87 0.58