Table 14.

Confidence intervals of the medians of Precision, Recall, F1-Score, and Accuracy for 250, 500, 750, and 1000 iterations of bootstrap with 90%.

Metric	Iterations	Std. Dev. of Median Differences	Confidence Interval	Significance
Precision	250	0.10	[−0.1, 0.26]	Non-Significant
	500	0.09	[−0.06, 0.23]	Non-Significant
	750	0.09	[−0.06, 0.23]	Non-Significant
	1000	0.09	[−0.06, 0.23]	Non-Significant
Recall	250	0.16	[0.07, 0.57]	Significant
	500	0.16	[0.07, 0.57]	Significant
	750	0.16	[0.07, 0.57]	Significant
	1000	0.15	[0.07, 0.57]	Significant
F1-Score	250	0.11	[0.05, 0.39]	Significant
	500	0.11	[0.03, 0.39]	Significant
	750	0.11	[0.03, 0.38]	Significant
	1000	0.11	[0.03, 0.38]	Significant
Accuracy	250	0.08	[−0.05, 0.24]	Non-Significant
	500	0.08	[0, 0.25]	Non-Significant
	750	0.08	[0, 0.24]	Non-Significant
	1000	0.08	[0, 0.24]	Non-Significant