A mathematical model and forecast for the coronavirus disease COVID-19 in Ukraine (Мc)

5 min readDec 13, 2020

(A.B. Alyokhin, B.V. Burkynskyi, A.N. Grabovoi, V.A, Dilenko, N.I. Khumarova)

Forecast Monitor

13/12/2020. No F-281.

https://www.facebook.com/ab.alyokhin/posts/217328673243369

This forecast devoted to the COVID-19 epidemic development in Ukraine for the period from December 14 to 20 (the main short-term forecast) and to January 10, 2021 (an approximate medium-term 4-week forecast) was compiled basis on official statistics available on December 13, using a set of development models of epidemics.

Compiling this forecast, a systemic epidemic model of the SEIRD type, SARIMA, ETS (Holt-Winters’ seasonal method), TCM, FB Prophet models and the author’s method of statistical modeling of time series with a seasonal component were used.

Currently, in the statistics (we emphasize this) of the COVID-19 epidemic development in Ukraine, there are obvious short-term positive shifts. For forecasting this presents certain difficulties associated with the contradiction between long-term and short-term trends in the development of the epidemic. This is especially evident using different forecasting methods and models, each (each) of which is looking for its own balance between these groups of trends. As a result, the use of such models leads to the generation of significantly different forecasts, to the formation of a wide forecast corridor.

In such circumstances, it is tempting to help models find the “right” solution, for example, by changing the parts of the history where models trained and tested. Modern models also have a wide range of parameters, varying which can also “make” the model behave “correctly”. However, all this is nothing more than the use of models as a tool for fitting forecasts to the expectations of developers.

The most appropriate forecasting tool in such conditions is the scenario approach. However, its use presupposes a thorough scientific study of scenarios, which is usually impossible without the participation of professionally trained representatives of consumers of such forecasts. In the event of forecasting the COVID-19 epidemic in Ukraine, the fulfillment of this condition is fundamentally impossible, which turns the scenario-forecasting tool into another toy in the hands of forecast developers, capable of generating forecasts for every taste.

Considering this, making forecasts, we do not allow the active influence of the intuition developers on the forecasting result, allowing the methods used manifest capabilities inherent in them. In particular, all models use statistical data from the entire observation period.

The results of the forecast calculations updated taking into account the circumstances noted above presented in Fig. 1–12 with corresponding indices.

In all charts, the “i” in the figure numbers indicates the interval forecast data, and the “c” index indicates the consensus forecast indicators. The interval forecast based on the lowest and highest values of particular forecasts; the consensus forecast is an averaging of particular forecasts.

Fig. 1–4 displayed:

- actual trajectories of the main cumulative indicators of the COVID-19 epidemic in Ukraine for the entire observation period;

- calculated trajectories of these indicators for a 28-day forecast period.

Diagrams Fig. 5–8 reflect:

- actual trajectories of the main daily indicators of the COVID-19 epidemic in Ukraine for the entire observation period.

- calculated trajectories of these indicators for a 28-day forecast period.

Fig. 9–12 are given:

- actual trajectories of the main synthetic indicators of the epidemic (the average absolute increase in the infected, the mortality rates I (TC) and (I (CC), as well as the progress indicator IP) for the entire observation period;

- estimated trajectory of these indicators for a 28-day (4-week) forecasting period.

As follows from the diagrams above, a number of forecast models take short-term trends as a basis, suggesting optimistic scenarios for the development of the epidemiological situation in the country. A number of models are wary of emerging new trends, only slightly correcting long-term trends in a more positive direction.

The situation is roughly reminiscent of the one in which the pathfinders, armed with flashlights at night, direct their light to the unknown object of interest to them, being in different positions relative to it. Where the light spots from different lights intersect, the subject better lit. However, the less illuminated details of this object are also very useful for its identification.

The situation is roughly reminiscent of the one in which the pathfinders, armed with flashlights at night, direct their light to the unknown object of interest to them, being in different positions relative to it. Where the light spots from different lights intersect, the subject better lit. However, for its more accurate identification, the less illuminated details of this object are also very useful.

The forecasting results generally indicate that it is still premature to talk about a turning point, even if it is insignificant, in the trends of the COVID-19 epidemic in Ukraine. This can judged only when this process becomes more forecastable. This evidenced by our long-term positive experience in practical forecasting of the COVID-19 epidemic in various countries of the world.

It will not be superfluous to recall that the long-term component of forecasts of the COVID-19 epidemic development must treated with the understanding that long-term forecasts indicate the possible scale of the epidemic, assuming that the trends existing at the time of making the forecasts will not change. Since the development of the epidemic influenced by the actions of the authorities and the actions of citizens, in conditions of inconsistent and poorly thought-out actions, such trends cannot remain changed for any length of time.

Sources of statistical data:

https://www.worldometers.info/coronavirus/#countries

https://covid19.rnbo.gov.ua/

Our materials also:

https://www.facebook.com/MATHMODELCOVID19

https://t.me/mathmodelcovid19

Accuracy of our forecasts:

https://www.facebook.com/ab.alyokhin/posts/154698732839697 (Germany)

https://www.facebook.com/ab.alyokhin/posts/142548897388014 (Spain)

https://www.facebook.com/ab.alyokhin/posts/150095069966730 (Italy)

https://www.facebook.com/ab.alyokhin/posts/148450556797848 (USA)

https://www.facebook.com/ab.alyokhin/posts/154364292873141 (Ukraine)