Denoising on COVID-19 t=300
0.1615MSE (sigma=0.5)I-RP-GFRFT
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| I-RP-GFRFTRotation Matrix Type=roll, k=5/5/52025.11 | 0.1615 | 0.3719 | 0.5401 | |
| FFT soft threshold2025.11 | 1.1262 | 1.1262 | 1.1262 | |
| 1D total variation2025.11 | 2.1405 | 5.1568 | 16.24 | |
| Tikhonov 1st diff.2025.11 | 3.6046 | 8.537 | 18.154 | |
| Median filter2025.11 | 3.6365 | 13.508 | 27.391 | |
| Moving average2025.11 | 3.8445 | 8.0091 | 18.889 | |
| Savitzky–Golay2025.11 | 4.7512 | 15.518 | 21.633 | |
| Gaussian smoothing2025.11 | 5.7046 | 14.672 | 22.372 | |
| GraphSAGE2025.11 | 139.7 | 507.63 | 1,139.1 | |
| GCN2025.11 | 141.02 | 512.03 | 1,103.3 | |
| ChebNet2025.11 | 141.26 | 489.9 | 1,095.5 | |
| GCNII2025.11 | 143.96 | 516.38 | 1,128 | |
| GAT2025.11 | 145.92 | 528.93 | 1,101.8 | |
| APPNP2025.11 | 152.01 | 498.72 | 1,195.6 |