Time-series Extrapolation on PhysioNet (test) using MNLL
0.32NMSE (Overall)DLFM
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| DLFMQ (Number of latent forces)=22021.06 | 0.32 | — | — | — | 0.88 | |
| DLFMQ (Number of latent forces)=12021.06 | 0.36 | — | — | — | 0.9 | |
| DLFMQ (Number of latent forces)=12021.06 | 0.39 | — | — | — | 0.75 | |
| DLFMQ (Number of latent forces)=22021.06 | 0.42 | — | — | — | 0.77 | |
| DGPKernel=EQ2021.06 | 0.44 | — | — | — | 1 | |
| DGPKernel=EQ2021.06 | 0.73 | — | — | — | 1.21 | |
| LFM-RFFLatent forces=2, Random features=202021.06 | 1 | — | — | — | — | |
| DGPKernel=ARC2021.06 | 1.01 | — | — | — | 2.6 | |
| VAR-GP2021.06 | 1.01 | — | — | — | 26.9 | |
| LFM-RFFLatent forces=2, Random features=202021.06 | 1.02 | — | — | — | — | |
| DNNArchitecture=single hidden layer of dimensionality 3002021.06 | 1.04 | — | — | — | — | |
| VAR-GP2021.06 | 1.33 | — | — | — | 15.5 | |
| DGPKernel=ARC2021.06 | 1.36 | — | — | — | 3.25 | |
| DNNArchitecture=single hidden layer of dimensionality 3002021.06 | 2.41 | — | — | — | — | |
| Conv-CNP2021.06 | — | 1.59 | 1.47 | 1.84 | — | |
| DLFMhidden layers=1, latent forces=1, random Fourier features=1002021.06 | — | 1.53 | 1.61 | 1.51 | — | |
| GNP2021.06 | — | 1.5 | 1.5 | 1.66 | — |