Intelligent Epidemiological Models for COVID-19

Abstract

The coronavirus COVID-19 is affecting around the world with strong differences between countries and regions. Extensive datasets are available for visual inspection and downloading. The material has limitations for phenomenological modelling but data-based methodologies can be used. This research focuses on intelligent modelling on the basis of these datasets. The methodology has been tested in the analysis of daily new confirmed COVID-19 cases and deaths in six countries. The datasets are studied per million people to get comparable indicators. Nonlinear scaling brings the data of different countries to the same scale and linear interactions represent the varying operating conditions well. The same approach operates for both the confirmed cases and deaths and can be used for any country or group of people. The effects of the vaccinations were clearly shown at the end of the analyzed period. During the pandemic, the scaling functions expanded for the confirmed cases but remained practically unchanged for the confirmed deaths which is consistent with increasing testing. Limitations are seen if there are too many interacting things, e.g. several infection transmission chains which are in different stages. The feasibility analysis needs to be extended to the modelling with inputs. The presented approach is promising for this wider analysis.