Table 3.

Evaluating the performance of several machine learning models including Support Vector Machine (SVM), Lasso Regression, Ridge Regression, Random Forest (RF), and XGBoost, for various treatments.

Treatment	Criteria	SVM	Lasso	Ridge	RF	XGBoost
cDMARDs	Sensitivity	0.946	0.944	0.946	0.955	0.955
	Specificity	0.806	0.830	0.789	0.860	0.852
	Precision	0.923	0.932	0.917	0.944	0.941
	Accuracy	0.906	0.911	0.900	0.927	0.925
Biologics	Sensitivity	0.946	0.947	0.955	0.953	0.948
	Specificity	0.814	0.831	0.807	0.868	0.852
	Precision	0.935	0.940	0.933	0.953	0.947
	Accuracy	0.912	0.917	0.916	0.931	0.923
Combination	Sensitivity	0.956	0.962	0.958	0.959	0.952
	Specificity	0.814	0.840	0.809	0.851	0.856
	Precision	0.934	0.944	0.933	0.947	0.948
	Accuracy	0.918	0.930	0.918	0.930	0.927
ADA	Sensitivity	0.960	0.965	0.971	0.971	0.965
	Specificity	0.631	0.754	0.646	0.846	0.800
	Precision	0.874	0.913	0.880	0.944	0.928
	Accuracy	0.870	0.908	0.882	0.937	0.920
ETA	Sensitivity	0.929	0.943	0.943	0.943	0.936
	Specificity	0.769	0.769	0.769	0.872	0.769
	Precision	0.936	0.937	0.937	0.964	0.936
	Accuracy	0.894	0.906	0.906	0.928	0.900
INF	Sensitivity	0.902	0.964	0.973	0.973	0.955
	Specificity	0.463	0.870	0.833	0.889	0.852
	Precision	0.777	0.939	0.924	0.948	0.930
	Accuracy	0.759	0.934	0.928	0.946	0.922
RIT	Sensitivity	0.909	0.938	0.931	0.925	0.934
	Specificity	0.875	0.890	0.846	0.897	0.890
	Precision	0.945	0.952	0.934	0.955	0.952
	Accuracy	0.899	0.923	0.906	0.917	0.921
TOC	Sensitivity	0.968	0.963	0.974	0.970	0.967
	Specificity	0.667	0.732	0.699	0.756	0.764
	Precision	0.931	0.943	0.938	0.949	0.950
	Accuracy	0.915	0.922	0.925	0.932	0.931
ABA	Sensitivity	0.857	0.888	0.882	0.901	0.907
	Specificity	0.861	0.851	0.832	0.931	0.881
	Precision	0.908	0.905	0.893	0.954	0.924
	Accuracy	0.859	0.874	0.863	0.912	0.897