Table 3.

Mean performance and standard deviations of KNE, KNU and DESKNN under different k_max values with 70-30 division.

Metrics	Methods	NON	kmax=2	kmax=3	kmax =4	kmax =5
Accuracy	KNE	0.9640±0.0123	0.9880±0.0060	0.9926±0.0067	0.9976±0.0042	0.9981±0.0023
	KNU	0.9551±0.0105	0.9728±0.0079	0.9896±0.0093	0.9952±0.0050	0.9904±0.0050
	DESKNN	0.9664±0.0122	0.9791±0.0076	0.9896±0.0073	0.9936±0.0051	0.9960±0.0033
F1	KNE	0.9710±0.0093	0.9897±0.0055	0.9934±0.0059	0.9981±0.0033	0.9986±0.0018
	KNU	0.9639±0.0082	0.9769±0.0070	0.9911±0.0075	0.9961±0.0045	0.9920±0.0051
	DESKNN	0.9723±0.0098	0.9814±0.0070	0.9915±0.0059	0.9948±0.0041	0.9968±0.0031
G-mean	KNE	0.9551±0.0165	0.9864±0.0061	0.9905±0.0080	0.9968±0.0035	0.9978±0.0027
	KNU	0.9442±0.0116	0.9693±0.0078	0.9892±0.0085	0.9957±0.0054	0.9921±0.0050
	DESKNN	0.9625±0.0147	0.9793±0.0071	0.9865±0.0087	0.9924±0.0050	0.9951±0.0035
AUC	KNE	0.9560±0.0159	0.9865±0.0061	0.9905±0.0080	0.9968±0.0035	0.9981±0.0022
	KNU	0.9551±0.0114	0.9728±0.0078	0.9896±0.0085	0.9952±0.0049	0.9904±0.0063
	DESKNN	0.9664±0.0145	0.9791±0.0070	0.9896±0.0081	0.9936±0.0046	0.9960±0.0032