23.05.2020. No 80 Forecast monitor
A mathematical model and forecast for the coronavirus disease COVID-19 in France
(A.B. Alyokhin, B.V. Burkynskyi, A.B. Brutman, A.I. Laiko, Z.N. Sokolovska)
Today’s release of the forecast monitor is devoted to the analysis of the of 10-day forecast accuracy for the COVID-19 outbreak in France, which was compiled according to statistics as of May 12th, 2020 inclusive, and published on May 13th, 2020.
The results of the forecast calculations and analysis and their degree of compliance with statistical data are described by tables and graphs Fig. 1, 2.
Assessment of forecast accuracy is given in table.1 and 2 (Fig. 1, 2
In table. 1 the calculated and actual values of the main indicators of coronavirus outbreak in France for 10-day forecast period are indicated, absolute and relative errors of daily forecasts and mean absolute error MAE and mean absolute percentage error MAPE as a whole
Table. 2 shows the same data for the case fatality rates I(TC) and I(CC), as well as the progress indicator IP
The data contained in these tables, including estimates of error, indicate a very high level of forecast accuracy.
Fig. 3–6 indicate as follows:
– actual trajectories of the main indicators of the coronavirus outbreak in France;
– calculated trajectory of these indicators;
– confidence intervals limits (ranges of possible deviations of the 10-days point forecast with a significance level of p=0.01).
As can be seen from graphs Fig. 3–6, all indicators of the outbreak were slightly deviated from forecasted values, indicating the outbreak’s trends stability in this country and high quality of forecasts.
Graphs Fig. 7–8, demonstrating the calculated trajectories for the longer term along with the actual trajectories, show that a sufficiently high accuracy of both long-term forecast and forecasted final values of cumulative reported cases and deaths should be expected at such level of 10-day forecast’s accuracy.
The graphs in Fig. 9–12 reflect as follows:
– actual trajectories of the indicators’ daily changes of the outbreak in France;
– calculated trajectory of these indicators throughout the entire period of observation, including the forecast period;
– confidence intervals limits (p=0,3), corresponding to the 10 — day forecast (for the period of May 13–22, 2020).
Let’s note, that the daily indicators are extremely difficult objects for forecasts, because of their great variability, due to the complexity of the forecasted processes. This explains the wide confidence intervals of the forecasts of these indicators, and a relatively high probability of error of the forecasts.
As can be seen, the actual values of daily indicators with a margin remained within confidence intervals and fluctuated around model trends.
Fig.13–15 reflect the following parameters of the coronavirus outbreak in France:
- actual trajectory of the fatality rates I(TC) and I(CC) and IP progress indicator for the entire observed period (until May 22, 2020 inclusive), and including 10-day forecast period;
- calculated trajectories of the indicators for this period.
It follows from the graphs fig.13–15 that the accuracy of reproduction by the model of the actual values of case fatality rates and indicator of progress as well as forecast accuracy of above stated indicators is quite high
Based on the results of the accuracy assessment for a 10-day forecast period (May 13–22, 2020) on May 12, 2020, the observed deviations of the actual data from the calculated ones, it can be concluded that the epidemiological situation in France is steadily and predictably developing towards the completion. However, within the framework of the rolling continuous forecasting concept, this cannot be the ground for refusal of new 10-day and 30-day’s forecast calculations, taking into account the new statistical data accumulated over the last 10 days (from May 13 to May, 22, 2020).
Source of statistics:
Our initiative group and mission:
Our publications on case fatality rates and indicator of progress: