A stable iterative method for refining discriminative gene clusters

BMC Genomics

Table 2 The performance of the algorithm for different classification studies

Datasets	θ _test		ρ_S(i) based on $\bar{d} (A_{i}, {Δ^{'}}_{2})$		ρ_all(i) based on $\hat{δ}$ (A_i)		$\bar{ω}$ (A_i)
Leukemia ALL T/B cell	0.970	0.977	1.72E-02	8.28E-03	0.088	0.331	0.406	0.973
Breast	0.843	0.842	1.33E-02	8.36E-03	0.142	0.421	0.351	0.974
Carcinoma	0.983	0.981	2.96E-02	3.20E-02	0.194	0.252	0.382	0.966
Colon	0.814	0.806	2.43E-02	2.06E-02	0.750	0.755	0.673	0.978
DLBCL	0.896	0.929	8.75E-02	1.99E-02	0.441	0.514	0.716	0.982
Melanoma	0.913	0.921	1.71E-02	2.25E-02	0.129	0.463	0.272	0.957
Prostate	0.889	0.916	4.79E-02	2.27E-02	0.495	0.541	0.680	0.987
SRBCT-BL	1.000	1.000	3.63E-04	7.52E-05	0.314	0.322	0.682	0.984
SRBCT-EWS	0.956	0.986	5.06E-04	9.17E-05	0.297	0.408	0.634	0.984
SRBCT-NB	0.989	0.996	2.99E-04	6.42E-05	0.321	0.436	0.665	0.986
SRBCT-RMS	0.974	0.980	4.82E-04	8.18E-05	0.304	0.347	0.630	0.989
Lukemia AML/ALL	0.966	0.972	5.62E-03	2.38E-03	0.212	0.398	0.627	0.980

ISSN: 1471-2164