Effective hybrid feature selection using different bootstrap enhances cancers classification performance

Table 14 The proposed methods compared with the OneRAttributeEval 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	99.154	97.625	1.529	0.978	0.980	0.978	7000	2.021	1.779	0.991	0.000502	97.625
DNA CNV	65.367	64.059	1.308	0.543	0.557	0.542	5000	1.020	0.644	0.813	0.001261	64.059
Parkinson’s disease	85.627	78.744	6.883	0.741	0.670	0.673	200	0.150	0.038	0.764	0.012796	78.744
Dermatology diseases	79.448	77.080	2.368	0.563	0.633	0.578	15	0.120	0.0006	0.909	0.001517	77.080
BreastEW	96.641	92.095	4.546	0.919	0.912	0.914	17	0.105	0.002	0.962	0.001380	92.095
Naïve base classifier
RNA gene	99.917	93.631	6.286	0.940	0.912	0.916	7000	2.021	0.076	0.949	0.000920	93.631
DNA CNV	67.368	66.528	0.840	0.617	0.625	0.610	5000	1.020	0.195	0.850	0.001035	66.528
Parkinson’s disease	74.262	73.784	0.478	0.400	0.434	0.415	200	0.150	0.001	0.715	0.009473	73.784
Dermatology diseases	86.004	83.589	2.415	0.812	0.791	0.758	15	0.120	0.0009	0.958	0.001873	83.589
BreastEW	93.451	93.196	0.250	0.469	0.460	0.462	17	0.105	0.001	0.986	0.003089	93.196
K-nearest neighbors(KNN) classifier
RNA gene	99.723	99.627	0.096	0.998	0.996	0.997	7000	2.021	0.010	0.999	0.000036	99.627
DNA CNV	78.155	71.777	6.378	0.651	0.603	0.597	5000	1.020	0.009	0.842	0.000189	71.777
Parkinson’s disease	80.688	72.481	8.207	0.388	0.443	0.414	200	0.150	0.0001	0.623	0.002529	72.481
Dermatology diseases	87.431	84.700	2.731	0.787	0.787	0.780	15	0.120	0.002	0.963	0.003370	84.700
BreastEW	94.728	92.976	1.752	0.930	0.921	0.924	17	0.105	0.001	0.961	0.000953	92.976

ISSN: 1756-0381