Registration Number
0226U003302,
(0124U000660)
, R & D reports
Title
Methods and means of artificial intelligence for preventing the spread of tuberculosis in military time
popup.stage_title
Проведення симуляцій на міĸро та маĸро рівнях.
Head
Vyklyuk Yaroslav I., Доктор технічних наук
Registration Date
26-03-2026
Organization
Lviv Polytechnic National University
popup.description1
The project aims to develop simulator programs based on multi-agent systems and neural networks to create a geospatial model of tuberculosis transmission and predict its spread. Additionally, it involves the development of methods to prevent the epidemic in the European region of the WHO.
popup.description2
Were achieved software simulators based on multi-agent systems were developed for modelling the transmission of tuberculosis infection. Simulation results of tuberculosis spread were obtained with visualization on a geographic map at different scales. Specifically, a mechanism for spatio-temporal visualization of modelling results using geoinformation data was implemented. Variants of tuberculosis prevention strategies were proposed and modelled. In particular, a set of alternative tuberculosis prevention and control strategies was developed, including mass screening, vaccination, isolation of infected individuals, increased treatment coverage, and combined approaches. A reinforcement learning neural network was developed for the automatic selection of an optimal tuberculosis prevention strategy. Specifically, a reinforcement learning model was implemented in which a control agent interacts with a multi-agent simulation environment. The novelty of the approach proposed in this study lies in its comprehensive nature and its focus on solving this class of problems.
The key advantage of the obtained results is the integrated combination of multi-agent modelling, geographic visualization, and reinforcement learning within a single simulation and analytical system adapted to real regional data. This ensures the practical value and international competitiveness of the results.
Product Description
popup.authors
Nataliya Boyko
Olena Vovk
Nataliya Melnykova
Denis Nevinsky
Oleksii Shamuratov
Dmytro Martianov
popup.nrat_date
2026-03-26