Predicting the Spread of COVID-19 in Riau Province Using the Recurrent Neural Network Method

Siringo ringo, Ivan Daniel
Hanif, Muhammad Abi
Agustina, Riska
Sari, Putri
Fitri, Adilla Anisa
Fauzi, Rifky

Indonesian Journal of Applied Mathematics and Statistics
Vol. 1 No. 1 (2024): Indonesian Journal of Applied Mathematics and Statistics (IdJAMS)

The COVID-19 pandemic has significantly impacted the world, including Riau Province, Indonesia. Predicting case developments accurately is crucial for controlling the virus's transmission. The purpose of this study is to forecast COVID-19 cases in the province of Riau by employing a Recurrent Neural Network (RNN). Between March 2020 and August 2021, data on daily cases, deaths, and recoveries was collected.  After testing the RNN model with a range of look-back values, the ideal values were found to be 20 days for daily cases and 10 days for daily recoveries and deaths. The model's ability to precisely capture the trend of real data was demonstrated by how closely it matched. Strong predictive performance was indicated by the resulting RMSE values, which were 435.31 for daily cases, 13.61 for daily deaths, and 331.95 for daily recoveries.