Machine learning-based models for the prediction of breast cancer recurrence risk

Table 2 Comparison of the prediction results of each test model using test datasets

Algorithms	AUC		Sensitivity	Specificity	PPV	NPV	F1 Score
AdaBoost	0.987	0.971	0.947	0.976	0.900	0.988	0.923
Decision Tree	0.894	0.951	0.941	0.953	0.800	0.988	0.865
GaussianNB	0.945	0.883	0.667	0.949	0.800	0.904	0.727
GBDT	0.967	0.971	0.947	0.976	0.900	0.988	0.923
LightGBM	0.983	0.971	0.947	0.976	0.900	0.988	0.923
LR	0.951	0.961	0.864	0.988	0.950	0.964	0.905
MLP	0.952	0.951	0.857	0.976	0.900	0.964	0.878
Random Forest	0.981	0.981	1.000	0.976	0.900	1.000	0.947
SVC	0.834	0.864	0.750	0.879	0.450	0.964	0.563
XGBoost	0.974	0.971	0.947	0.976	0.900	0.988	0.923
LDA	0.847	0.883	0.722	0.918	0.650	0.940	0.684

Abbreviations: PPV, positive predictive value; NPV, negative predictive value; LR, logistic regression; RF, random forest; SVC, support vector classification; XGBoost, extreme gradient boosting; GBDT, gradient boosting decision tree; MLP, multilayer perceptron; LDA, linear discriminant analysis; AdaBoost, adaptive boosting; GaussianNB, Gaussian naive Bayes; LightGBM, light gradient boosting machine

ISSN: 1472-6947