BMC Bioinformatics

Table 3 Comparison of machine MLs with and without the proposed FFS model based on accuracy

From: Fractal feature selection model for enhancing high-dimensional biological problems

Dataset	FFS parameters						Accuracy (%)
	n	p	NB		DT		RF		SVM		KNN
	n	p	ML	FFS’	ML	FFS’	ML	FFS’	ML	FFS’	ML	FFS’
ALLAML	2	16	89.03	91.25	86.67	91.25	90.48	100	73.33	89.03	89.03	89.03
COLON	10	21	43.77	52.15	69.23	80.18	84.21	94.74	61.54	73.2	74.54	94.74
Lung_discrete	4	81	66.68	75.84	66.68	77.24	76.19	95.24	66.68	66.68	86.67	95.24
Lung	2	36	75.37	80.74	75.61	88.04	81.82	90.91	70.73	72.68	72.68	90.91
Lymphoma	10	16	45.39	55.1	60.18	70	70.28	90.00	55.82	67.28	65.00	80
TOX_171	2	86	74.29	86.25	57.14	66.43	75	96.15	34.29	44.89	74.29	96.15
WarpPIE10P	10	6	90.24	93.65	83.33	83.22	76.19	93.65	52.86	63.65	88.62	93.65
Orlraws10P	10	21	80.99	91.45	61.44	77.14	79.31	100	50	63.05	91.45	91.45
CLL_SUB_111	2	26	69.56	86.97	65.22	76.32	69.57	86.97	34.78	45.22	65.22	86.97
GLI_85	4	36	70.59	71.28	76.47	96.55	79.31	96.55	64.71	78.24	88.24	88.24

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com