Table 8:

Gaussian data with n = 200 and p = 251 under MNAR. Approximately 40% features and 90% cases contain missing values.

Models	Style	Time(s)	Imp MSE	Bias(${\widehat{\beta }}_1$)	CR(${\widehat{\beta }}_1$)	SE(${\widehat{\beta }}_1$)	SD(${\widehat{\beta }}_1$)
SoftImpute	SI	15.3	0.0194	−0.0997	0.84	0.1182	0.1358
GAIN	SI	53.2	0.8618	0.6212	0.18	0.1502	0.5088
MIWAE	SI	47.6	0.0502	0.0695	0.90	0.1356	0.1410
not-MIWAE	SI	41.7	1.4701	0.1040	0.65	0.1084	0.1624
Linear RR	SI	3009.6	0.0658	0.1823	0.90	0.1782	0.0935
MICE	MI	48.6	0.0233	−0.0049	0.93	0.1160	0.1244
Sinkhorn	MI	29.9	0.0757	0.0117	0.97	0.1839	0.1523
MI-NNGP1	MI	3.4	0.0116	0.0069	0.93	0.1147	0.1215
MI-NNGP1-BS	MI	3.4	0.0149	0.0140	0.96	0.1285	0.1179
MI-NNGP2	MI	10.4	0.0085	−0.0024	0.95	0.1123	0.1148
MI-NNGP2-BS	MI	10.3	0.0094	−0.0018	0.96	0.1177	0.1147
Complete data	-	-	-	−0.0027	0.90	0.1098	0.1141
Complete case	-	-	-	0.2518	0.89	0.3385	0.3319
ColMean Imp	SI	-	0.1414	0.3539	0.72	0.2210	0.1712