Using Markov Chains to Predicate Pandemic Trend: A Case Study in Libya for COVID–19.

Abstract

Many predictive models have been developed by various academic institutions to support health systems in strategic decision-making, planning, and policies that help in the challenge against COVID-19. These models are useful in determining, the expected number of cases and deaths due to COVID–19, as well as the required resources such as hospital beds for isolation period and ICU, and necessary supplies such as protective equipment. In this article, the stationary Markov Chain is applied to the Libyan population to predict the status of the pandemic in Libya after more than four years of its spread. The data used was collected by WHO, and the results showed that the chain had converged due to the large sample size taken, resulting in the limiting probability being very close to the initial distribution. Additionally, the probability of staying in a good situation is 70.9% and to become worse is 29.1%. Finally, due to the convergence of the chain, these results will remain the same regardless of the initial state of the chain.