Machine learning approaches for the genomic prediction of rheumatoid arthritis and systemic lupus erythematosus

Table 1 Comparison of machine learning model performance with 5-fold cross-validation

Classifier	Accuracy	Precision	Sensitivity	Specificity	F1 score	AUC
Logistic Regression	0.7610	0.7385	0.7801	0.7430	0.7587	0.8451
Random Forest	0.9402	0.9376	0.9384	0.9420	0.9379	0.9871
Support Vector Machine	0.9373	0.9310	0.9398	0.9352	0.9353	0.9829
Gradient Tree Boosting	0.9635	0.9579	0.9668	0.9606	0.9623	0.9953
Extreme Gradient Boosting	0.9618	0.9544	0.9668	0.9573	0.9606	0.9948

RA: rheumatoid arthritis; SLE: systemic lupus erythematosus; AUC: area under the curve

ISSN: 1756-0381