Arthritis Research & Therapy

Table 4 Model comparison of three machine learning methods in the third phase

From: Establishment of a differential diagnosis method and an online prediction platform for AOSD and sepsis based on gradient boosting decision trees algorithm

	RF		GBDT		LR
	Test set	Validation set	Test set	Validation set	Test set	Validation set
AUC	0.9222	0.9832	0.9222	0.9916	0.8333	0.9229
ACC	0.8947	0.9171	0.8947	0.9457	0.7895	0.8229
Sens	0.8889	0.9667	0.8889	0.9556	0.7778	0.8814
Spec	0.9000	0.8737	0.9000	0.9578	0.8000	0.8050
MCC	0.7889	0.8422	0.7889	0.8981	0.5778	0.6864

Abbreviations: AUC area under curve, ACC accuracy, Sens sensitivity, Spec specificity, MCC Matthews correlation coefficient

Back to article page

ISSN: 1478-6362

Contact us

General enquiries: journalsubmissions@springernature.com