Assessing the risk of high-grade squamous intraepithelial lesions (HSIL+) in women with LSIL biopsies: a machine learning-based study

Table 5 Predictive performance of different machine learning models in validation set

Model	ACC (%)	AUC	Sensitivity	Specificity	PPV	NPV
Logistic regression	72.1	0.843	0.600	0.969	0.750	0.941
Naïve bayes	79.5	0.814	0.700	0.809	0.359	0.946
SVM	93.4	0.801	0.600	0.985	0.857	0.942
KNN	92.7	0.861	0.500	0.992	0.909	0.929
Decision tree	94.7	0.936	0.800	0.969	0.800	0.969
Random forest	90.7	0.940	0.800	0.924	0.615	0.968
XGBoost	88.1	0.893	0.800	0.893	0.533	0.967

ISSN: 1750-9378