Table 21.

Performance comparison of our method vs. CMIM for the imbalanced E. coli data set when identical numbers of features are selected.

	AUC		Precision		Recall		F1		MCC
N4	0.691 >	0.663	0.725	0.717	0.280	0.271	0.404	0.393	0.391	0.381
N5	0.690	0.686	0.737	0.710	0.295 >	0.264	0.421 >	0.385	0.407 >	0.373
N6	0.701	0.697	0.742	0.715	0.287 >	0.265	0.414 >	0.387	0.403 >	0.376
N7	0.714	0.693	0.735	0.711	0.275 >	0.261	0.400 >	0.382	0.392 >	0.371
N8	0.705	0.690	0.742	0.709	0.288 >	0.254	0.415 >	0.373	0.405 >	0.364
N9	0.707	0.701	0.726	0.720	0.293 >	0.271	0.417 >	0.394	0.401 >	0.382
N10	0.711	0.702	0.724	0.692	0.294 >	0.248	0.418 >	0.364	0.401 >	0.353
N11	0.714	0.698	0.732	0.690	0.278 >	0.247	0.403 >	0.363	0.393 >	0.351
N12	0.712	0.690	0.725	0.683	0.292 >	0.239	0.416 >	0.353	0.400 >	0.342
N13	0.714	0.688	0.733	0.678	0.287 >	0.236	0.413 >	0.349	0.400 >	0.337