A mathematical model and forecast for the coronavirus disease COVID-19 in the World
(A.B. Alyokhin, B.V. Burkynskyi, A.N. Grabovoi, V.A, Dilenko, N.I. Khumarova)
08/10/2020. No F-225.
This monitor contains estimates of the forecast accuracy of the COVID-19 Pandemic development in the world, compiled with a lead of 10 days for the period from September 27 to October 6, 2020 (https://www.facebook.com/ab.alyokhin/ posts / 184308869878683). When making this forecast, used statistical data from September 26, 2020. The forecast accuracy estimated using statistical data inclusive October 6, 2020.
An important statement.
This concludes the forecasting of the COVID-19 Pandemic. This is our second and probably the last time. Our decision is caused by the extremely unsatisfactory state of the Pandemic database (DB) of our main source (https://www.worldometers.info/coronavirus/worldwide-graphs/).
The data in the database changed next day for the previous days. Summarizing the results, each time we adjust our database following the changes in the Source. As a result, forecasts based on one data, and checked on the other. However, that is not all. Since the data in the database continues retroactively change, the subsequent reevaluation of the estimates of the accuracy of our forecasts will be different. This absurdity can only stopped by the most absurd activity — forecasting a Pandemic based on low quality data.
Thus, the Pandemic, as an object of forecasting, sent to the basket, where the COVID-19 epidemic is already waiting for it. In the USA, it is the most likely source of unpredictability of statistical data in general for all countries in the world.
The results of assessing the accuracy of the forecast and additional illustrations to them given in table. 1–2 and fig. 1–13.
Table 1 shows the calculated and actual values of the main indicators of the COVID-19 pandemic for the entire forecast period. The absolute and relative errors of daily forecasts, as well as the average absolute MAE error and the average absolute error in percent MAPE *, including the same MAPE * estimate on an accrual basis, forecast in the whole.
Table 2 shows the same data for case fatality rates I (TC), I (CC) and the IP progress indicator.
As can see from the data in table. 1 the forecast accuracy is (as of today) high. Mean absolute percentage errors (MAPE * estimates) for the main pandemic indicators are at the following level:
- total number of infected people — 0.18%;
- total number of deaths — 0.55%;
- total number of recovered — 0.45%.
The accuracy of predicting the derived indicators of the coronavirus pandemic in the world (MAPE * estimates) is as follows (see Table 2):
- mortality rate I (TC) — 0.70%;
- mortality rate I (CC) — 0.26%;
- IP progress indicator — 0.54%,
Due to the high degree of coincidence of actual and forecast indicators in Fig. 1–4, the boundaries of the confidence intervals hidden due to their lack of information.
Let us also recall that the trajectories of the actual values of the pandemic parameters are showing in red, and the forecast ones — in blue.
In Fig. 1–4 indicate:
- actual trajectories of the main indicators of the COVID-19 pandemic for the entire observation period, including the forecast period;
- calculated trajectories of these indicators for the forecast period.
Diagrams 5–8 characterize:
- actual trajectories of changes in daily pandemic indicators for the entire observation period, including the 10-day forecast period;
- calculated trajectories of these indicators for the forecast period;
- confidence intervals of the forecast with 95% confidence level.
Diagrams Fig. 9–13 reflect such synthetic parameters of the pandemic as:
- actual trajectories of the statistical reproductive number SR0, the average absolute increase in the number of infected, mortality rates I (TC) and I (CC), as well as the IP progress indicator for the entire observation period, including the 10-day forecast period;
- calculated trajectories of the indicated indicators for the forecast period.
As follows from the tables and diagrams, the rate of coronavirus spread in the world remains practically unchanged (Fig. 5–9), which ensures a constant growth of the main cumulative indicators (Fig. 1–4, 10). In addition, even the record outbursts of some of them (Fig. 6 and Fig. 7) cannot affect the general patterns of the pandemic development.
The dynamics of the mortality rate I (TC) is still negative (Fig. 11) and is due to the continued rapid growth in the number of infected. The growth of the IP progress indicator is solely due to the excess of the growth rate of the number of closed cases over the growth rate of the number of infected, but not due to the decrease in the increase in the number of infected, which is a sign of the transition of the pandemic to the completion stage.
The high accuracy of the forecast indicates the long-term stability of negative trends in the development of the COVID-19 pandemic and the inability of most countries in the world to make positive adjustments to them.
Sources of statistical data:
Our materials also:
Accuracy of our forecasts:
Publications on case fatality rates and progress indicator