Table 3.

Classification and explainability metrics formulas.

Ensemble trees algorithm	Description
Accuracy	${\displaystyle \frac{(TP+TN)}{(TP+TN+FP+FN)}(1)}$
Sensitivity/Recall	${\displaystyle \frac{TP}{(TP+FN)}(2)}$
Specificity	${\displaystyle \frac{TN}{(TN+FP)}(3)}$
Balanced accuracy	${\displaystyle \frac{Sensivity+Specificity}{2}(4)}$
Precision	${\displaystyle \frac{TP}{(TP+FP)}(5)}$
F1-Score	${\displaystyle 2\ast \frac{Precision\ast Sensivity}{Precision+Sensivity}(6)}$
Interpretability (I)	${\displaystyle \frac{maskedfeatures}{totalinputfeatures}(7)}$
Interpretability-accuracy index (IAI)	$I\ast Acc(9)$
Interpretability-concordance index (ICI)	$I\ast Cindex(9)$

(TN as true negative, FN as false negative, FP as false positive, and TP as true positive)