Table 2. Overall summary of performance of algorithms.
Models | Population A | Population B | |||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
TPR | FPR | Precision | Recall | F | MCC | AUC | TPR | FPR | Precision | Recall | F | MCC | AUC | ||
BN | 0.887 | 0.133 | 0.888 | 0.887 | 0.887 | 0.745 | 0.93 | 0.801 | 0.255 | 0.799 | 0.801 | 0.798 | 0.568 | 0.858 | |
RF | 0.858 | 0.237 | 0.858 | 0.858 | 0.853 | 0.668 | 0.94 | 0.773 | 0.354 | 0.799 | 0.773 | 0.751 | 0.515 | 0.86 | |
SL | 0.922 | 0.127 | 0.922 | 0.922 | 0.921 | 0.82 | 0.975 | 0.73 | 0.395 | 0.737 | 0.73 | 0.707 | 0.402 | 0.714 | |
LMT | 0.922 | 0.127 | 0.922 | 0.922 | 0.921 | 0.82 | 0.975 | 0.773 | 0.302 | 0.771 | 0.773 | 0.766 | 0.503 | 0.762 |
BN means Bayes Net, RF means Random Forest, SL denotes Simple Logistic. F means F measure. MCC means Matthews correlation coefficient.