Fuzzy-Based Prediction of Spread of Covid-19 Pandemic

Abstract

Necessity of containment of the spread of contagious virus cannot be overemphasized, COVID-19 inclusive. COVID-19 outbreak has shown how globally interconnected we are. Effects of its spread on individuals, communities, and nations in terms of trade and development are awesome and importance of prevention (containment) cannot be downplayed. Different methods employed in prediction of the spread of COVID-19 spread include linear regression model, machine learning technique, multiplicative calculus, Google trends, etc. Cost of prevention is much cheaper than containment and inherent prowess in soft computing paradigms, hence the design and implementation of predictive Fuzzy Inference System (FIS) model for the spread of COVID-19. The design employed Mandanin FIS for prediction of the spread of COVID-19 virus. Eight (8) COVID-19 symptoms were employed as inputs into the design which lead to three options at the output. The output are: “not infected”, “quarantine”, and “infected”. Triangular and trapezoidal membership functions were employed for fuzzification of inputs into the FIS while trapezoidal was employed at the output stage. The fuzzified 8 inputs were appropriately employed in formation of the fuzzy rule system of the FIS model. 128 group of fuzzy inference rule system was employed in the implementation. Performance of the FIS model revealed a great performance of the predictive system.