Short- and long-term predictions of novel corona virus using mathematical modeling and artificial intelligence methods

Alkhammash, H. I.; Al Otaibi, S.; Ullah, N.

This paper proposes spread prediction of novel corona virus outbreak using different compartmental models and artificial intelligence (AI) methods. Real data for several months is collected from the Ministry of Health (MOH) website, Kingdom of Saudi Arabia and two compartmental models, namely SIR (susceptible, infectious, recovered) and SEIRD (susceptible, exposed, infectious, recovered, dead) are utilized to best fit the data. AI methods are well suited for short- and long-term stochastic forecasts. Keeping in view the inherent advantages of AI methods, adaptive neuro-fuzzy inference system (ANFIS) models are trained using the collected data to replicate the dynamic behavior of the COVID-19 spread in Kingdom of Saudi Arabia. The prediction comparison for COVID-19 spread is made between the compartmental and ANFIS models for both short- and long-term forecasts of the experimental data. From the presented results, ANFIS-based models show superior performance as compared to compartmental models.