Table 4 Scenario 4: Single binary incomplete variable missing at random, single continuous complete variable, 100 random runs, 50% missing rate
From: Secure distributed multiple imputation enables missing data inference for private data proprietors
Final analysis | Imputation | Performance | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
Technology | Θ bias | Θ SD | Θ rMSE | \(| \widehat{{\rm{y}}}-{\rm{y}}|\) (μ) | \(| \widehat{{\rm{y}}}-{\rm{y}}|\) (σ) | Discr. | Accuracy | AUC | Time (s) | Net (MB) | |
500 inds. | Python | 0.378 | 0.0 | 0.378 | 0.19 | 0.026 | 0 | 0.636 | 0.466 | 0.024 | N/A |
PyMICE | 0.010 | 2.48 × 10-16 | 0.010 | 0.019 | 0.015 | N/A | 0.632 | 0.631 | 0.047 | N/A | |
SMC-MICE | 0.012 | 3.85 × 10-3 | 0.012 | 0.023 | 0.015 | 0 | 0.632 | 0.719 | 0.231 | 8.359 | |
MHE-MICE | 0.028 | 2.50 × 10-3 | 0.028 | 0.042 | 0.026 | 0 | 0.640 | 0.719 | 470.4 | 8298 | |
5000 inds. | Python | 0.348 | 2.48 × 10-16 | 0.348 | 0.184 | 0.121 | 0 | 0.529 | 0.510 | 0.081 | N/A |
PyMICE | 0.004 | 0.0 | 0.004 | 0.024 | 0.014 | N/A | 0.525 | 0.500 | 0.031 | N/A | |
SMC-MICE | 0.060 | 1.08 × 10-1 | 0.123 | 0.105 | 0.045 | 0 | 0.518 | 0.467 | 2.016 | 86.63 | |
MHE-MICE | 0.082 | 4.97 × 10-3 | 0.083 | 0.087 | 0.038 | 0 | 0.523 | 0.467 | 1737 | 60,778 | |
