Novel COVID-19 Prediction Model in Python Using FB Prophet

Abstract

Since end of 2019, the coronavirus disease has spread quickly to nations all over the world. This has had a huge impact on many nations' economies and also on the global health system. As there is no effective treatment for cure, it's indeed vital to foresee COVID situations ahead to make the appropriate plans. Despite the fact that there are many models for predicting COVID-19, none of them have predicted for a specific number of days i.e., 30 days or 90 days or 1 week. To address this issue, the proposed study employs the Facebook prophet model for the job of COVID-19 case predictions using Python over the following thirty days. The prophet is a data-driven time series projecting technique using an incremental approach that matches non-linear tendencies with annual, monthly, and everyday periodicity and vacation impacts.