The COVID-19 pandemic in Ukraine: regional context and short-term forecast

On May 9, scientists from the World Data Center "Geoinformatics and Sustainable Development" at Igor Sikorsky Kyiv Polytechnic Institute presented new results of modeling the peculiarities of the spread of the coronavirus pandemic in the regions of Ukraine. At the same time, significant territorial unevenness of this process across the country, differences in communications and religious traditions, uneven migration flows, regional peculiarities of countering and combating the disease, etc. were taken into account. Therefore, the study is called: "Foresight COVID-19: regional context" (

It was carried out using methods of technical analysis of time series based on the indicators "zigzag" and "supertrend". Traditionally, these indicators are used to track the main trends in the behavior of stock markets, which, by the nature of their volatility, are similar to the behavior of pandemic processes.

The authors of the study analyzed the progress of the spread of coronavirus disease in all regions of Ukraine and in Kyiv. It turned out that the maximum rate of change in the number of cases of COVID-19 in Ukraine reached on the 51st-52nd day from the first registered patient, i.e. April 22, 2020 (April 16 in Kyiv). After reaching it, due to the application of quarantine measures, a tendency to decrease in the number of newly identified patients per day was shown. However, in most regions this process has more or less volatility, i.e. it is too early to talk about a steady downward trend in the number of new coronavirus cases. Moreover, there is not enough data for analysis in some regions (Donetsk, Luhansk, Mykolaiv, Sumy, Kherson, Khmelnytsky and Chernihiv regions).

The researchers also provided a short-term forecast for the development of the pandemic (, which was made using a multi-layer neural network Back Propagation based on the "sliding window" mechanism. According to it, for Ukraine in a ten-day period of time (from May 6 to May 15) the nature of the development of a pandemic can probably get a clear linear character with a "plateau" principle of 400-550 new infected per day. This nature of the development of the process can last until the third decade of May 2020, during which a pandemic peak is likely to occur. After this, the pandemic can probably decline (the number of people infected daily will be steadily lower than the number of those who recover daily).

As for Kyiv, during the ten-day period (May 6 – May 15, 2020) the processes of pandemic development will probably still more non-stationary with quite high volatility.