Table 12.

Two-sided paired-samples t test comparing the performance of ISMB with other algorithms. For each performance measure the number on top is the mean difference between ISMB and the indicated algorithm and the number at the bottom is the corresponding p-value. The mean difference is negative when ISMB has a lower score on a performance measure than the competing algorithm. On all measures, a negative mean difference indicates better performance by ISMB. Underlined results indicate p-values of 0.05 or smaller.

Performance measure	ISMB-MS	NISMB	NB	DT	LR	NN	kNN	LBR	AB
Misclassification error	−0.004	−0.001	−0.021	−0.013	−0.012	−0.019	−0.005	−0.007	−0.003
	0.077	0.312	0.014	0.021	0.065	<0.001	0.334	0.289	0.258
AUC	−0.001	−0.002	−0.001	−0.104	−0.017	−0.032	−0.023	−0.001	−0.020
	0.077	0.242	0.975	<0.001	0.022	<0.001	<0.001	0.932	0.001
Logarithmic loss	−0.009	−0.004	−0.163	−0.211	−0.215	−0.306	−0.140	−0.071	−0.103
	0.001	0.026	0.005	<0.001	0.006	<0.001	<0.001	<0.001	0.002
Squared error	−0.004	−0.001	−0.044	−0.044	−0.034	−0.062	−0.023	−0.019	−0.016
	0.003	0.054	0.002	<0.001	0.009	0.002	<0.001	0.017	0.007
CAL score	−0.003	−0.002	−0.033	−0.011	−0.003	−0.047	−0.008	−0.016	−0.002
	0.044	0.058	<0.001	0.018	0.441	<0.001	0.079	0.001	0.237