Imputation on PhysioNet Challenge 2012 (test)
0.186MAESAITS
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SAITS# of parameters=5.32M, s/epoch=52022.02 | 0.186 | 26.6 | 0.431 | |
| Transformer# of parameters=4.36M, s/epoch=3.12022.02 | 0.19 | 26.9 | 0.445 | |
| SAITS-base# of parameters=1.38M, s/epoch=2.72022.02 | 0.192 | 27.3 | 0.439 | |
| BRITS# of parameters=0.73M, s/epoch=12.82022.02 | 0.256 | 36.5 | 0.767 | |
| BRITSModel Type=RNN2018.05 | 0.281 | 39.14 | — | |
| RITSModel Type=RNN2018.05 | 0.3 | 41.89 | — | |
| BRITS-IModel Type=RNN2018.05 | 0.361 | 50.01 | — | |
| ImputeTSModel Type=Non-RNN2018.05 | 0.39 | 54.2 | — | |
| RITS-IModel Type=RNN2018.05 | 0.395 | 54.8 | — | |
| GP-VAE# of parameters=0.15M, s/epoch=40.12022.02 | 0.398 | 56.7 | 0.63 | |
| M-RNNModel Type=RNN2018.05 | 0.451 | 62.65 | — | |
| M-RNN# of parameters=0.07M, s/epoch=6.82022.02 | 0.533 | 76 | 0.776 | |
| GRU-DModel Type=RNN2018.05 | 0.559 | 77.58 | — | |
| MFModel Type=Non-RNN2018.05 | 0.622 | 87.68 | — | |
| MICEModel Type=Non-RNN2018.05 | 0.634 | 89.17 | — | |
| E2GAN# of parameters=0.08M, s/epoch=22.82022.02 | 0.702 | 100.1 | 0.964 | |
| MeanModel Type=Non-RNN2018.05 | 0.72 | 100 | — | |
| Median2022.02 | 0.726 | 103.5 | 0.988 | |
| KNNModel Type=Non-RNN2018.05 | 0.732 | 101.66 | — | |
| GRUI-GAN# of parameters=0.16M, s/epoch=14.42022.02 | 0.765 | 109.1 | 1.04 | |
| Last2022.02 | 0.862 | 123 | 1.207 |