Table 5.

Performance of the consensus and the combined models on the test set.

Consensus Model Based on a Set of Rules
Error significance ε ¹		Coverage	ACC	MCC	CCR	Sens	Spec	NPV	PPV
0.05		0.51	0.86	0.72	0.86	0.84	0.88	0.88	0.84
0.10		0.71	0.79	0.59	0.80	0.83	0.77	0.88	0.70
0.20		0.89	0.77	0.54	0.78	0.82	0.73	0.85	0.68
0.30		0.83	0.78	0.56	0.79	0.85	0.72	0.88	0.68
Combined Model Based on Mean p-Values
Error significance ε	Validity	Efficiency	ACC	MCC	CCR	Sens	Spec	NPV	PPV
0.05	0.97	0.24	0.89	0.77	0.89	0.92	0.86	0.94	0.82
0.10	0.93	0.46	0.86	0.72	0.87	0.94	0.80	0.95	0.76
0.20	0.82	0.79	0.77	0.56	0.78	0.85	0.72	0.87	0.69
0.30	0.71	0.95	0.75	0.50	0.76	0.80	0.72	0.84	0.66

¹ Error significance of the underlying model, not of the combined model itself.