Table 3 Scenario 3: Single continuous 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. | \(| \widehat{{\rm{y}}}-{\rm{y}}|\) (μ) | \(| \widehat{{\rm{y}}}-{\rm{y}}|\) (σ) | Time (s) | Net (MB) | |
500 inds. | Python | 0.014 | 5.24 × 10-4 | 0.014 | 0.028 | 0.019 | 0 | 0.392 | 0.238 | 0.023 | N/A |
PyMICE | 0.014 | 2.22 × 10-16 | 0.014 | 0.028 | 0.018 | N/A | 0.393 | 0.235 | 0.046 | N/A | |
SMC-MICE | 0.013 | 2.35 × 10-3 | 0.013 | 0.028 | 0.018 | 0 | 0.392 | 0.238 | 0.038 | 1.686 | |
MHE-MICE | 0.016 | 2.40 × 10-3 | 0.016 | 0.031 | 0.021 | 0 | 0.388 | 0.248 | 285.9 | 9,629 | |
5000 inds. | Python | 0.002 | 3.09 × 10-4 | 0.002 | 0.021 | 0.013 | 0 | 0.494 | 0.074 | 0.086 | N/A |
PyMICE | 0.015 | 0.0 | 0.015 | 0.021 | 0.014 | N/A | 0.494 | 0.072 | 0.031 | N/A | |
SMC-MICE | 0.002 | 4.06 × 10-2 | 0.040 | 0.040 | 0.019 | 0 | 0.494 | 0.074 | 0.206 | 16.176 | |
MHE-MICE | 0.069 | 2.40 × 10-2 | 0.069 | 0.051 | 0.027 | 0 | 0.494 | 0.087 | 1910 | 86,628 | |
