A parallel approach to the fractional time delay model for predicting the spread of COVID-19

Due to the importance of forecast accuracy for diseases such as COVID-19, the existence of a mathematical model is particularly important. In this research, first, a model to describe the spread of the COVID-19 pandemic is examined. This model is based on a fractional ordinary differential equation. Then the predictor-corrector numerical method is presented to solve this model. Due to the computational challenge of numerically solving fractional models, a task-parallel approach with coarse granularity is presented to solve this model on shared memory systems. The initial data for testing the proposed approach is the data reported on December 31, 2019 by the Wuhan Municipal Commission of the outbreak of the COVID-19 pandemic in the city of Wuhan, China. The numerical results obtained from the proposed parallel approach show that the speedup of the parallel method compared to the sequential method reaches 2.76 in the prediction of 1000 days.