Predicting the dynamics of the coronavirus (COVID-19) epidemic based on the case-based reasoning approach

Abstract

The case-based rate reasoning (CBRR) method is presented for predicting future values of the coronavirus epidemic’s main parameters in Russia, which makes it possible to build shortterm forecasts based on analogues of the percentage growth dynamics in other countries. A new heuristic method for estimating the duration of the transition process of the percentage increase between specified levels is described, taking into account information about the dynamics of epidemiological processes in countries of the spreading chain. The CBRR software module has been developed in the MATLAB environment, which implements the proposed approach and intelligent proprietary algorithms for constructing trajectories of predicted epidemic indicators.