Table 2.

Accuracy statistics, including sample size (N), the number of presences (N[p]), prevalence (prev), calculated by (N[p]/N), area under the curve (AUC), the true skills statistic (TSS), and standard deviation (SD)

Data set	N	N(p)	Prev	AUC	TSS	Sensitivity	SD	Specificity	SD
Training	299	218	0.73	0.9	0.63	0.79	0.02	0.84	0.04
2000	72	51	0.71	0.90	0.58	0.96	0.03	0.62	0.12
1990	136	78	0.57	0.8	0.49	0.95	0.03	0.41	0.07
1980	354	209	0.59	0.8	0.33	0.91	0.02	0.38	0.04
1970	113	81	0.72	0.6	0.14	0.93	0.03	0.22	0.07