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

Table 7

Experimental result analysis of different classifiers for with and without PCA on dataset-2 (CKD-21).

Classifier	Recall		TNR		PPV		F1-score		Train accuracy		Test accuracy		ROC-AUC
	With PCA	Without PCA	With PCA	Without PCA	With PCA	Without PCA	With PCA	Without PCA	With PCA	Without PCA	With PCA	Without PCA	With PCA	Without PCA
ADAB	100	100	100	100	100	100	100	100	100	100	100	100	1	1
DT	100	97	100	97	100	97	100	97	100	100	100	97.4	1	0.974
ELM	100	53	100	56	100	54	100	53	98	100	100	54.55	0.729	0.53
GB	100	97	100	95	100	95	100	96	99.44	98.88	100	96.1	1	0.961
KNN	100	76	100	95	100	94	100	84	96.65	86.59	100	85.71	1	0.856
SVM	100	76	100	38	100	55	100	64	97.77	61.45	100	57.14	1	0.574
RF	100	100	100	97	100	97	100	98	99.44	100	100	98.7	1	0.99
LR	97	100	97	92	97	93	97	96	98.32	96.09	97.4	96.1	0.995	0.99
NB	97	84	97	100	97	100	97	91	96.65	93.85	97.4	92.21	0.993	0.99
SGD	100	100	95	0	95	49	97	66	98.32	50.28	97.4	49.35	1	0.86
MLP	92	92	100	97	100	97	96	93	100	89.39	96.1	94.81	0.987	0.95
PAC	95	100	95	0	95	49	95	66	98.32	50.28	94.81	49.35	0.987	0.5

The best results are indicated in bold.