Table 1 The performance of the following six multi-classification ML models in the test cohort (n = 96).

		LR	SVM	RFC	LGBM	Catboost	ANN
	Accuracy (%)	68.8	67.7	76.0	74.0	76.0	65.6
	Kappa	0.486	0.483	0.571	0.534	0.594	0.456
	Weighted_Precision	0.739	0.758	0.757	0.737	0.788	0.745
	Weighted-Recall	0.688	0.677	0.760	0.740	0.760	0.656
	Weighted-F1	0.702	0.698	0.758	0.736	0.770	0.678
	AUC	0.901	0.883	0.904	0.899	0.912	0.898
None	Precision	0.936	0.951	0.879	0.877	0.940	0.950
	Recall	0.772	0.684	0.895	0.877	0.825	0.667
	F1	0.846	0.796	0.887	0.877	0.879	0.784
	AUC	0.722	0.781	0.806	0.809	0.809	0.737
Mild	Precision	0.455	0.441	0.609	0.556	0.571	0.419
	Recall	0.435	0.652	0.609	0.652	0.696	0.565
	F1	0.444	0.526	0.609	0.600	0.627	0.481
	AUC	0.867	0.865	0.851	0.838	0.884	0.887
Moderate and severe	Precision	0.444	0.524	0.533	0.500	0.556	0.480
	Recall	0.750	0.688	0.500	0.375	0.625	0.750
	F1	0.558	0.595	0.516	0.429	0.588	0.585

ML, machine learning; LR, logistic regression; SVM, support vector machine; LGBM, light gradient boosting machine; Catboost, categorical boosting; ANN, artificial neural network.

Quick links

Search