Effective hybrid feature selection using different bootstrap enhances cancers classification performance

Table 16 The proposed methods compared with the PCA method

Datasets	Train Data %	Test Data %	Over-fitting Diff. %	Pre	Rec	F1-score	NO.F	F-Time (sec)	C-Time (sec)	AUC	Var.	ACC %
J48 classifier
RNA gene	96.823	94.884	1.939	0.943	0.954	0.942	700	1.027	0.174	0.985	0.000740	94.884
DNA CNV	62.334	60.665	1.669	0.526	0.516	0.505	2800	49.795	1.549	0.827	0.001724	60.665
Parkinson’s disease	83.745	73.816	9.929	0.625	0.607	0.602	250	0.079	0.028	0.645	0.002418	73.816
Dermatology diseases	81.148	80.878	0.270	0.576	0.663	0.604	18	0.016	0.003	0.925	0.000142	80.878
BreastEW	95.723	93.493	2.230	0.942	0.925	0.929	20	0.016	0.002	0.959	0.000832	93.493
Naïve base classifier
RNA gene	87.072	79.403	7.669	0.794	0.806	0.794	700	1.027	0.005	0.954	0.002116	79.403
DNA CNV	29.336	27.641	1.695	0.255	0.352	0.233	2800	49.795	0.077	0.680	0.000319	27.641
Parkinson’s disease	74.471	73.821	0.65	0.604	0.558	0.545	250	0.079	0.009	0.698	0.002826	73.821
Dermatology diseases	98.361	96.179	2.182	0.961	0.952	0.953	18	0.016	0.0001	0.997	0.001025	96.179
BreastEW	90.041	89.803	0.238	0.896	0.886	0.889	20	0.016	3.057	0.962	0.001707	89.803
K-nearest neighbors(KNN) classifier
RNA gene	99.750	99.740	0.010	0.999	0.997	0.998	700	1.027	0.002	0.999	0.000059	99.740
DNA CNV	81.200	74.348	6.852	0.663	0.639	0.634	2800	49.795	0.010	0.867	0.000273	74.348
Parkinson’s disease	81.158	72.612	8.546	0.612	0.571	0.575	250	0.079	0.001	0.627	0.002308	72.612
Dermatology diseases	92.380	87.162	5.218	0.862	0.860	0.844	18	0.050	0.00002	0.969	0.001984	87.162
BreastEW	94.728	92.976	1.752	0.930	0.921	0.924	20	0.030	0.0003	0.961	0.000953	92.976

ISSN: 1756-0381