Dynamic Bayesian networks in solving the tasks of strategic and operational planning to hold back the pandemic

The use of government measures to hold back the pandemic requires effective solution of operational and strategic planning tasks and coverage of the entire population of the country without exception. This study aims to identify the interrelationships between the indicators of the epidemic process and the indicators of the state's response to hold back the COVID-19 pandemic. A sequence of anti-epidemic measures has been identified that contributes to reducing the incidence of the population and increasing the economic potential of the country. The study was carried out on the example of Russia. Weekly statistical data for the period from January to December 2020 were used on indicators characterizing the development of the COVID-19 epidemic process and the implementation of state anti–epidemic measures (source - Oxford COVID-19 Government Response Tracker). A dynamic Bayesian network is constructed. When training the network, the algorithm "a variation on Ghada Trabelsi's dynamic max-min hill climbing" was used. The calculations were performed using the dbnR library of the R programming language. The following types of relationships have been identified between indicators of the development of the COVID-19 epidemic process and government measures to contain the pandemic: 1) short-term; 2) long-term; 3) self-influence. State anti-epidemic measures form a complex dynamic structure that determines the patterns of their impact on curbing the increase in the incidence of COVID-19 in the Russian population. The epidemic process is characterized by a memory effect: changes in the morbidity and mortality of the population at the current moment will cause changes in the indicators of this process in the future. Effective implementation of state anti-epidemic measures helps to reduce the increase in morbidity and mortality from COVID-19. At the same time, it is necessary to observe a strict sequence of anti–epidemic measures: start with the implementation of measures in the field of healthcare, then gradually connect restrictive state measures, then - measures of state economic support for the population. Special attention should be paid to the implementation of two types of activities: 1) restrictions on holding meetings; 2) economic support for the population, as they are significantly influenced by various factors. At the same time, the restriction on the movement of public transport does not have a significant impact on the morbidity and mortality of the population from COVID-19.

1. Doubled mortality rate during the COVID-19 pandemic in Italy: quantifying what is not captured by surveillance / A. Odone, D. Delmonte, G. Gaetti, C. Signorelli // Public Health. 2021. Vol. 190. Pp. 108–115. URL: https://doi.org/10.1016/j.puhe.2020.11.016 (date of access: 14.01.2024).

2. Yousefpour A., Jahanshahi H., Bekiros S., Optimal policies for control of the novel coronavirus disease (COVID-19) outbreak // Chaos, Solitons & Fractals. 2020. Vol. 136. URL: https://doi.org/10.1016/j.chaos.2020.109883 (date of access: 14.01.2024).

3. The impact of various policy factors implemented for controlling the spread of COVID-19 / P.N. Senthil Prakash, B. Hariharan, S. Kaliraj [et al.] // Materials Today : proceedings. 2021. URL: https://doi.org/10.1016/j.matpr.2021.01.524 (date of access: 14.01.2024).

4. Wang Y. Government policies, national culture and social distancing during the first wave of the COVID19 pandemic: International evidence // Safety Science. 2021. Vol. 135. URL: https://doi.org/10.1016/j.ssci.2020.105138 (date of access: 14.01.2024).

5. COVID-ABS: An agent-based model of COVID-19 epidemic to simulate health and economic effects of social distancing interventions / Petrônio C.L. Silva, Paulo V.C. Batista, Hélder S. Lima [et al.] // Chaos, Solitons & Fractals. 2020. Vol. 139. URL: https://doi.org/10.1016/j.chaos.2020.110088 (date of access: 14.01.2024).

6. Public health effects of travel-related policies on the COVID-19 pandemic: A mixed-methods systematic review / Lama Bou-Karroum, Joanne Khabsa, Mathilda Jabbour [et al.] // Journal of Infection. 2021. Vol. 83, Issue 4. Pp. 413–423. URL: https://doi.org/10.1016/j.jinf.2021.07.017 (date of access: 14.01.2024).

7. López D.Q., Castilla G.A.V. Gaussian dynamic Bayesian networks structure learning and inference in R. 2019. URL: https://www.researchgate.net/publication/337545669_Gaussian_dynamic_Bayesian_networks_structure_learning_and_inference_in_R (date of access: 14.01.2024).

8. Trabelsi G. New structure learning algorithms and evaluation methods for large dynamic Bayesian networks. URL: https://tel.archives-ouvertes.fr/tel-00996061/document (date of access: 14.01.2024).

9. Ontañón S., Montaña J.L., Gonzalez A.J. A Dynamic-Bayesian Network framework for modeling and evaluating learning from observation // Expert Systems with Applications. 2014. Vol. 41, Issue 11. Pp. 5212– 5226. URL: https://doi.org/10.1016/j.eswa.2014.02.049 (date of access: 14.01.2024).