Extreme learning machine versus classical feedforward network

Table 21 p-values from Nemenyi test for regularisation coefficient C, for F1-measure

Regularisation coefficient	0.0001	0.0003	0.001	0.003	0.01	0.03	0.1	0.3	1	3	10	30	100	300
0.0001
0.0003	0.900
0.001	0.325	0.900
0.003	0.022	0.497	0.900
0.01	0.002	0.140	0.900	0.900
0.03	0.001	0.016	0.550	0.900	0.900
0.1	0.001	0.014	0.523	0.900	0.900	0.900
0.3	0.001	0.071	0.838	0.900	0.900	0.900	0.900
1	0.019	0.470	0.900	0.900	0.900	0.900	0.900	0.900
3	0.382	0.900	0.900	0.900	0.900	0.497	0.470	0.785	0.900
10	0.838	0.900	0.900	0.890	0.497	0.114	0.101	0.325	0.864	0.900
30	0.900	0.900	0.890	0.252	0.049	0.004	0.003	0.022	0.231	0.900	0.900
100	0.900	0.900	0.353	0.025	0.002	0.001	0.001	0.001	0.022	0.411	0.864	0.900
300	0.900	0.707	0.038	0.001	0.001	0.001	0.001	0.001	0.001	0.049	0.300	0.900	0.900
Average ranking	11.000	9.235	7.235	5.824	4.941	3.941	3.882	4.588	5.765	7.353	8.353	9.765	10.941	12.176

The p-values that are below the assumed significance level are expressed in bold