Table 3 Performance metrics of machine learning models across feature sets

Set	Model	Accuracy	Sensitivity	Specificity	PPV	NPV	AUC
EDA (6 features)	LogisticRegression	0.848	0.808	0.852	0.317	0.981	0.901
	RandomForest	0.842	0.808	0.845	0.308	0.981	0.908
	LightGBM	0.849	0.769	0.856	0.313	0.978	0.898
	XGBoost	0.854	0.692	0.867	0.308	0.971	0.876
HighFrequency (12 features)	LogisticRegression	0.851	0.837	0.853	0.326	0.984	0.907
	RandomForest	0.847	0.798	0.851	0.313	0.980	0.908
	XGBoost	0.870	0.731	0.882	0.345	0.975	0.895
	LightGBM	0.862	0.712	0.875	0.326	0.973	0.896
Intersection (8 features)	LogisticRegression	0.854	0.788	0.859	0.323	0.979	0.905
	RandomForest	0.852	0.769	0.859	0.317	0.978	0.908
	XGBoost	0.864	0.740	0.875	0.335	0.975	0.886
	LightGBM	0.870	0.740	0.881	0.345	0.976	0.896
Minimalistic (4 features)	RandomForest	0.838	0.779	0.843	0.297	0.978	0.898
	LightGBM	0.839	0.740	0.847	0.292	0.975	0.880
	LogisticRegression	0.842	0.721	0.853	0.294	0.973	0.892
	XGBoost	0.845	0.712	0.857	0.297	0.972	0.873

Quick links

Search