Table 3.
Technique (reference) | NRMSEa | |
|
20% imputation performance | 30% imputation performance |
MICEb [77] | 0.8479 | 0.8886 |
SoftImpute [78] | 0.8546 | 0.9044 |
KNNc-based imputation [79] | 1.0209 | 1.0044 |
Deep autoencoder model [58] | 0.7926 | 0.8453 |
aNRMSE: normalized root mean squared error.
bMICE: multivariate imputation by chained equations.
cKNN: K-nearest neighbors.