Machine learning-based Diagnostic model for determining the etiology of pleural effusion using Age, ADA and LDH

Respiratory Research

Table 4 Area under the receiver operating characteristic curve of single etiologies in each machine learning method in the test set

	LR	SVM	XGBoost	RF	KNN	Tab Transformer
MPE	0.925	0.891	0.931	0.934	0.925	0.935
TPE	0.939	0.913	0.953	0.950	0.946	0.947
PPE	0.745	0.706	0.765	0.706	0.652	0.752
Transudative	0.966	0.946	0.967	0.969	0.899	0.975
Others	0.636	0.644	0.813	0.666	0.538	0.766

MPE Malignant pleural effusion, TPE Tuberculous pleural effusion, PPE Parapneumonic pleural effusion, Transudative Transudative pleural effusion, Others Other causes. LR multinomial linear regression, SVM support vector machine, XGBoost Extreme Gradient Boosting, RF random forest, KNN K-Nearest Neighbors, Tab Transformer Tabular Transformer

ISSN: 1465-993X