Table 1.

Statistical testing of NRF results for the 9x9 dataset, with regard to autoencoder (AE) and matrix factorization (MF). Statistically significant improvements achieved by PhyloMissForest (under any of the considered configuration profiles) are denoted as $✓$, while non-significant differences are marked with ×

Dataset	%Missing	Non-bootstrap p-values		Bootstrap p-values		PhyloMissForest diff.
		Vs. AE	Vs. MF	Vs. AE	Vs. MF	Vs. AE	Vs. MF
9x9	5%	0.09	0.25	0.02	0.10	$✓$	$✓$
	10%	0.00	0.02	0.00	0.01	$✓$	$✓$
	15%	0.02	0.08	0.01	0.09	$✓$	$✓$
	20%	0.02	0.02	0.11	0.14	$✓$	$✓$
	25%	0.25	0.28	0.22	0.12	×	×
	30%	0.00	0.14	0.00	0.12	$✓$	×
	35%	0.58	0.17	0.97	0.44	×	×
	40%	0.09	0.31	0.19	0.35	$✓$	×
	45%	0.31	0.44	0.06	0.10	$✓$	$✓$
	50%	0.14	0.17	0.09	0.53	$✓$	×
	55%	0.11	0.44	0.01	0.09	$✓$	$✓$
	60%	0.01	0.91	0.00	0.35	$✓$	×

Bold values refer to p-values denoting statistically significant improvements