An Intelligent Diagnostic System to Analyze Early-Stage Chronic Kidney Disease for Clinical Application

Table 12

Average performance analysis of ML models on all four types of dataset.

S/L	Classifier	Training accuracy (%)	Testing accuracy (%)	ROC-AUC
1	RF	99.77	98.48	0.98
2	ADAB	100	98.46	0.98
3	SVM	98.77	98.17	0.98
4	SGD	99.15	97.83	0.98
5	KNN	98.34	97.78	0.98
6	MLP	100	97.73	0.99
7	DT	100	97.7	0.91
8	GB	99.86	97.7	0.98
9	LR	98.86	97.7	0.99
10	ELM	98.5	97.56	0.99
11	PAC	98.96	96.95	0.98
12	NB	97.01	95.79	0.99

The best results are indicated in bold.