Skip to main content
[Preprint]. 2024 Mar 25:2024.01.13.24301248. Originally published 2024 Jan 15. [Version 2] doi: 10.1101/2024.01.13.24301248

Table 6:

Performance of different forecasting models during critical time periods in 1-week ahead

Model Persistence FIGI-Net CDC Karlen CovidAnalysis JHU MIT Microsoft
Slope Dissimilarity 0.316±0.079 0.127±0.076 0.273±0.108 0.5±0.147 0.57±0.136 0.269±0.134 0.606±0.166 0.421±0.137
RMSE 7015.59±15176.43 3964.97±6690.04 7773.23±14058.41 13056.12±14803.5 10394.45±21478.21 7241.44±9418.22 11170.42±22911.14 12775.19±26537.04
MAPE 19.96±3.7 11.15±2.97 24.48±27.51 49.08±76.09 46.45±23.76 21.07±57.56 42.64±29.46 33.64±47.7
P-value(v.s. FIGI-Net) 1.99×10−9 - 4.89×10−8 9.31×10−10 5.15×10−10 1.92×10−6 5.45×10−13 5.15×10−10