Table 6:

Exponential data with n = 200 and p = 1001 under MAR. Approximately 40% features and 92% cases contain missing values. Here Linear RR and MICE are not included due to running out of RAM.

Models	Style	Time(s)	Imp MSE	Bias(${\widehat{\beta }}_1$)	CR(${\widehat{\beta }}_1$)	SE(${\widehat{\beta }}_1$)	SD(${\widehat{\beta }}_1$)
SoftImpute	SI	37.8	0.6284	−0.5896	0.92	0.6348	0.4858
GAIN	SI	130.9	1.691	−0.7217	0.40	0.4611	2.229
MIWAE	SI	58.4	1.530	0.5626	0.05	0.1522	0.1456
Sinkhorn	MI	42.1	0.3845	−0.2077	1.00	0.4566	0.2832
MI-NNGP1	MI	3.1	0.3296	−0.0421	0.75	0.1757	0.3220
MI-NNGP1-BS	MI	3.4	0.4543	−0.0570	1.00	0.3366	0.2466
MI-NNGP2	MI	3.9	0.2358	0.1098	0.80	0.2312	0.3383
MI-NNGP2-BS	MI	10.3	0.2516	−0.0242	0.95	0.3203	0.3037
Complete data	-	-	-	0.0156	0.95	0.0978	0.0938
Complete case	-	-	-	0.1726	0.89	0.3984	0.4534
ColMean Imp	SI	-	0.3794	−0.1506	1.00	0.4556	0.3123