Table 4. Classification error rate, sensitivity and Brier score produced by Random Forest, k-Nearest Neighbors and Support Vector Machine classifiers on srbct dataset based on genes selected by the given methods. The best result is shown in bold.

		RF						kNN						SVM
Genes		POS	RPOS	GClust	sigF	Wilc	mRmR	POS	RPOS	GClust	sigF	Wilc	mRmR	POS	RPOS	GClust	sigF	Wilc	mRmR
5	Err	0.048	0.019	0.096	0.040	0.021	0.390	0.078	0.034	0.100	0.000	0.074	0.078	0.086	0.021	0.035	0.007	0.328	0.412
	BS	0.005	0.002	0.096	0.029	0.023	0.236	0.007	0.001	0.057	0.002	0.071	0.074	0.037	0.003	0.028	0.011	0.217	0.255
	sen	0.919	0.988	0.961	0.980	0.978	0.549	1.000	1.000	0.718	1.000	0.915	0.914	0.878	0.998	0.942	0.984	0.608	0.574
10	Err	0.018	0.021	0.027	0.035	0.013	0.086	0.039	0.038	0.055	0.000	0.071	0.069	0.016	0.011	0.029	0.006	0.204	0.143
	BS	0.002	0.003	0.029	0.027	0.022	0.089	0.004	0.002	0.041	0.000	0.076	0.071	0.016	0.002	0.031	0.013	0.138	0.093
	sen	0.999	0.991	0.957	0.981	0.977	0.879	1.000	1.000	0.852	1.000	0.925	0.918	0.992	0.995	0.943	0.998	0.766	0.785
15	Err	0.004	0.014	0.016	0.001	0.013	0.165	0.039	0.035	0.075	0.000	0.074	0.071	0.004	0.004	0.015	0.002	0.188	0.182
	BS	0.002	0.002	0.028	0.021	0.024	0.142	0.002	0.002	0.047	0.000	0.071	0.073	0.005	0.001	0.015	0.010	0.118	0.129
	sen	0.995	0.991	0.956	0.998	0.977	0.805	1.000	1.000	0.807	1.000	0.927	0.910	0.995	1.000	0.962	1.000	0.756	0.764
20	Err	0.009	0.007	0.016	0.010	0.009	0.081	0.036	0.036	0.053	0.000	0.066	0.071	0.011	0.003	0.020	0.002	0.144	0.130
	BS	0.002	0.002	0.029	0.021	0.023	0.088	0.002	0.002	0.041	0.000	0.069	0.072	0.007	0.001	0.019	0.010	0.098	0.082
	sen	0.987	0.990	0.956	1.000	0.986	0.875	1.000	1.000	0.895	1.000	0.919	0.911	0.997	0.999	0.986	1.000	0.797	0.816
25	Err	0.009	0.004	0.017	0.011	0.009	0.067	0.038	0.020	0.060	0.000	0.066	0.074	0.011	0.008	0.030	0.000	0.134	0.098
	BS	0.002	0.002	0.031	0.021	0.024	0.084	0.002	0.001	0.039	0.001	0.071	0.072	0.006	0.002	0.023	0.008	0.087	0.067
	sen	0.992	0.997	0.956	1.000	0.987	0.881	1.000	1.000	0.870	1.000	0.923	0.915	0.999	0.997	0.977	1.000	0.826	0.885
30	Err	0.006	0.006	0.023	0.007	0.005	0.075	0.034	0.002	0.047	0.000	0.064	0.065	0.009	0.014	0.018	0.000	0.131	0.129
	BS	0.002	0.002	0.029	0.022	0.024	0.094	0.002	0.001	0.040	0.001	0.069	0.070	0.006	0.002	0.017	0.006	0.087	0.090
	sen	0.992	0.997	0.957	1.000	0.994	0.883	1.000	1.000	0.866	1.000	0.914	0.924	0.998	0.999	0.951	1.000	0.828	0.855