Multivariate Time Series Imputation On 2
평가 지표
OOB Rate (10^−3)
Path Difference
Path Length
Player Distance
Step Change (10^−3)
평가 결과
이 벤치마크에서 각 모델의 성능 결과
모델 이름 | OOB Rate (10^−3) | Path Difference | Path Length | Player Distance | Step Change (10^−3) | Paper Title | Repository |
|---|---|---|---|---|---|---|---|
| BRITS (SingleRes) | 3.874 | 0.571 | 0.702 | 0.417 | 4.811 | BRITS: Bidirectional Recurrent Imputation for Time Series | |
| GRUI | 4.703 | 0.690 | 1.141 | 0.398 | 14.95 | Multivariate Time Series Imputation with Generative Adversarial Networks | - |
| NAOMI | 1.733 | 0.581 | 0.573 | 0.423 | 2.565 | NAOMI: Non-Autoregressive Multiresolution Sequence Imputation | |
| MaskGAN | 4.592 | 0.680 | 0.793 | 0.427 | 9.622 | MaskGAN: Better Text Generation via Filling in the______ | - |
| KNN | 0.128 | 0.746 | 0.921 | 0.403 | 13.24 | - | - |
0 of 5 row(s) selected.