Skip to main content
. 2020 Dec 30;6:212–221. doi: 10.1016/j.idm.2020.12.007

Table 3.

Fitting (a) and forecasting (b) performance statistics for the exponential decay model applied to WRt (Eq. (5)) and the Gompertz model applied to the CI (Eq. (7)). For each time interval, we highlight the lowest MAPE (gray) for the size of the error.

a) Model metrics
Model
N° Days
WR
CI
MAPE
rPers
MAPE
rPers
P1 - 17/03
20
0.056
0.968
0.036
1.000
P2 - 24/03
27
0.051
0.978
0.058
1.000
P3 - 31/03
34
0.044
0.986
0.141
1.000
P4 - 07/04
41
0.044
0.990
0.151
1.000



b) Post-model metrics (until 21 April 2020)
Model
N° Days until 21/04
WR
CI
MAPE
rPers
MAPE
rPers
P1 - 17/03 35 0.075 0.998 0.371 0.984
P2 - 24/03 28 0.029 0.997 0.322 0.996
P3 - 31/03 21 0.037 0.999 0.040 0.999
P4 - 07/04 14 0.023 1.000 0.028 0.999