Epidemic forecasting model developed by scientists from St Petersburg University proves its effectiveness
Scientists from St Petersburg University have confirmed the effectiveness of their previously developed mathematical model for forecasting epidemic development. The analysis was based on retrospective data from the coronavirus pandemic in Moscow and St Petersburg during 2020–2021. The probability of model error did not exceed 1%.
In 2021–2022, a team of scientists from the Centre for Dynamic Processes and System Analysis at St Petersburg University developed a new approach to studying dynamic inflow and outflow systems with stochastic parameters, along with a new methodology for forecasting their dynamics. Using this method, the University’s mathematicians were able to identify new peaks in the growth of COVID-19 incidence and key indicators of the disease’s spread.
The research findings are published in the scientific journal "Problems of virology".
The system was based on the hypothesis that the dynamics of disease incidence and epidemic spread are influenced by numerous natural factors. To model this, the mathematicians from St Petersburg University employed a dynamic game against nature as a mathematical framework. It was found that the spread of new viruses, as well as population growth dynamics in individual countries or even globally, can be described using a model with stochastic (random) parameters. Using this approach, the researchers were able to accurately forecast the increase in the number of cases in St Petersburg on several occasions.
The model developed by the mathematicians from St Petersburg University successfully forecasted the trajectory of infections at the peak of the pandemic. Victor Zakharov spoke about this in an interview with the Izvestia newspaper. After just one month of observation, the model was able to forecast the total number of infections for the following three to four weeks with high accuracy. Later, the mathematicians refined their model, adapting it to forecast the spread of any epidemic.
Mathematical models of dynamic systems must be continuously updated to account for an increasing number of factors. To validate the efficiency and accuracy of their model, the scientists conducted a retrospective analysis and generated real-time forecasts of virus spread, using COVID-19 in St Petersburg and Moscow at the peak of the pandemic as a case study.
"The retrospective two-week forecasts of the total number of cases and active COVID-19 cases presented in the study demonstrated high accuracy for both Moscow and St Petersburg. The MAPE (Mean Absolute Percentage Error) of the total number of cases at the peaks of incidence did not usually exceed 1%, which can be considered a good and reliable result from the perspective of epidemic forecasting," said Victor Zakharov, Academic Director of the Centre for Dynamic Processes and System Analysis and Professor at St Petersburg University.
As part of their study, the University scientists compared forecasted and actual figures for the number of new coronavirus cases, the total number of patients, and their dynamics in relation to the emergence of new COVID-19 strains in Moscow and St Petersburg. Forecasts made before April 2020 showed an average deviation of up to 20% from actual case numbers. By early May 2020, when the model had been trained on March—April data, the errors did not exceed 1% on average.
The study demonstrated that the model developed by the mathematicians from St Petersburg University outperformed the existing SIR and ARIMA models. Until recently, these models were the primary forecasting tools available to public authorities for epidemic monitoring and response. However, St Petersburg University has proposed a more accurate alternative.
Notably, the epidemic forecasting model project was supported by a grant from the St Petersburg Science Foundation, which was established in 2021 on the initiative of Governor Alexander Beglov.
St Petersburg University, the oldest university in Russia, was founded on 28 January (8 February) 1724. This is the day when Peter the Great issued a decree establishing the University and the Russian Academy of Sciences. St Petersburg University today is a major centre for education and research. More than 20,000 students study here, and more than 15 major laboratories and 23 resource centres have been established as part of the country’s leading Research Park. Graduates of the University have been recipients of the Nobel and Fields Prizes on multiple occasions.
Recently, St Petersburg, the Northern Capital of Russia, officially introduced a new holiday — Day of St Petersburg University — which has been included in the St Petersburg Law "On holidays and commemorative days in St Petersburg".
In February 2025, a ceremonial event was held, during which Roscosmos cosmonauts presented the University with the "300th anniversary of St Petersburg University" flag, which had travelled to the International Space Station and back.