Table 5.

Comparison of experimental results of classification accuracy (%), sensitivity (%), specificity (%), and AUC (in terms of mean ± standard deviation) using a RF classification model, where 𝛜 indicates 1.0 × 10⁻⁵.

Feature Set	Validation	Accuracy	Sensitivity	Specificity	AUC
FS1	5-fold	86.67 ± 1.56	76.58 ± 1.27	95.35 ± 2.68	0.8603 ± 0.0152
	10-fold	86.09 ± 1.59	77.03 ± 1.35	93.9 ± 2.31	0.8546 ± 0.0154
	Leave-one-out	85.78 ± 1.24	76.35 ± 1.79	93.9 ± 2.83	0.8512 ± 0.0115
FS2	5-fold	76.25 ± 2.28	63.97 ± 4.03	86.82 ± 2.89	0.7539 ± 0.0234
	10-fold	76.67 ± 1.38	65.76 ± 4.03	86.05 ± 2.68	0.7591 ± 0.0148
	Leave-one-out	76.75 ± 1.00	65.4 ± 2.02	86.51 ± 3.08	0.7596 ± 0.0087
FS3	5-fold	73.25 ± 0.61	75.68 ± 1.71	71.16 ± 1.14	0.7342 ± 0.0065
	10-fold	72.68 ± 1.45	75.68 ± 1.45	70.1 ± 1.94	0.7289 ± 0.0153
	Leave-one-out	72.50 ± 𝛜	75.68 ± 𝛜	69.77 ± 𝛜	0.7272 ± 𝛜
FS4	5-fold	73.75 ± 𝛜	51.35 ± 𝛜	93.02 ± 𝛜	0.7219 ± 𝛜
	10-fold	73.57 ± 0.44	50.96 ± 0.94	93.02 ± 𝛜	0.72 ± 0.0047
	Leave-one-out	73.75 ± 𝛜	51.35 ± 𝛜	93.02 ± 𝛜	0.7219 ± 𝛜
FS5	5-fold	84.82 ± 1.82	77.22 ± 1.97	91.36 ± 2.05	0.8429 ± 0.0182
	10-fold	87.32 ± 1.56	79.54 ± 1.97	94.02 ± 1.69	0.8678 ± 0.0157
	Leave-one-out	86.13 ± 1.42	77.30 ± 1.32	93.72 ± 2.09	0.8551 ± 0.0138
FS6	5-fold	83.75 ± 0.95	75.29 ± 1.73	91.03 ± 2.30	0.8316 ± 0.0089
	10-fold	84.58 ± 1.56	76.58 ± 3.12	91.47 ± 1.09	0.8402 ± 0.0165
	Leave-one-out	86.38 ± 1.42	78.65 ± 2.24	93.02 ± 1.80	0.8584 ± 0.0144
FS7	5-fold	84.86 ± 1.5	77.78 ± 2.47	90.96 ± 2.31	0.8437 ± 0.015
	10-fold	85.63 ± 1.53	77.67 ± 1.31	92.73 ± 2.44	0.8505 ± 0.0147
	Leave-one-out	86.25 ± 1.48	77.30 ± 1.32	93.95 ± 2.59	0.8564 ± 0.0141