Exploiting cellular automata and linear regression to predict disease spread

Abstract

The Covid-19 epidemic has significantly impacted the world, highlighting the urgent need to understand and anticipate its mechanisms of spread. This essential knowledge is necessary to plan and conduct an immediate and adequate public health response. This study presents an approach to modeling the spread of Covid-19 using non-uniform cellular automata. The paper extends the application of a previously developed model that uses cellular automata and the SIRD epidemiological model for predicting the evolution of Covid-19. Originally developed for a different context, the model is now adapted to assess the progression of the pandemic in Germany and Italy, considering the potential impact of neighboring countries on the spread of the epidemic. Additionally, this approach expands the prediction model to countries lacking infection data (Switzerland) by using estimated parameters from neighboring countries and randomly initialized parameters for the target country. The study demonstrates the model’s precision in tracking infection rates over time by employing reliable public data sources such as the World Health Organization and existing information from national health websites. The study not only furnishes valuable insights into the regional distribution of the epidemic’s impact but also makes a significant contribution by extending the model’s application beyond the borders of a single country. It introduces a strategy for extrapolating patterns of infection spread across borders, marking it as the first study of its kind with substantial importance in the field.