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

Table 4 Predictive performance of different machine learning models in training set

Model	ACC (%)	AUC	Sensitivity	Specificity	PPV	NPV
Logistic regression	79.1	0.768	0.561	0.835	0.395	0.908
Naïve bayes	88.2	0.741	0.509	0.882	0.453	0.903
SVM	89.0	0.722	0.456	0.973	0.765	0.903
KNN	80.5	0.764	0.544	0.855	0.419	0.907
Decision tree	87.6	0.849	0.596	0.929	0.618	0.923
Random forest	86.4	0.824	0.579	0.919	0.579	0.919
XGBoost	83.9	0.793	0.561	0.892	0.500	0.914

ISSN: 1750-9378