Manuscript Title:

COMPARISON OF PERFORMANCE FOR FORECASTING THE NEW CASES OF COVID-19 IN MALAYSIA BY USING ARIMA, LSTM AND PROPHET

Author:

MUHAMMAD SYAFIQ ALZA BIN ALIAS, NORAZLIN BINTI IBRAHIM, ZALHAN BIN MOHD ZIN

DOI Number:

DOI:10.17605/OSF.IO/NM3K8

Published : 2021-12-10

About the author(s)

1. MUHAMMAD SYAFIQ ALZA BIN ALIAS - Industrial Automation Section, UniKL Malaysia France Institute, Bangi, Malaysia.
2. NORAZLIN BINTI IBRAHIM - Industrial Automation Section, UniKL Malaysia France Institute, Bangi, Malaysia
3. ZALHAN BIN MOHD ZIN - Industrial Automation Section, UniKL Malaysia France Institute, Bangi, Malaysia

Full Text : PDF

Abstract

Currently, Malaysia are facing the third wave of COVID-19. The number of new cases is very alarming because this disease can bring death to affected people. Government can take early precautions to prevent the spiking of COVID-19 new cases if this disease can be detected early. Therefore, forecasting the new cases in the future can help to alert the government on the rising of COVID-19 new cases. Besides that, the proven data analysis of COVID-19 can also be used to further convince the people about the potential threat that might occur innear future. Hence, the people will know that the actions taken is not solely based on assumption, which in current situation everyone is vulnerable and pressured mainly due to the economic status. This research is conducted to compare and discover the most suitable time series forecasting model that can be used to predict the new cases of COVID-19 in Malaysia. The models used are ARIMA, LSTM and Prophet. The result shows that ARIMA model with (2,1,1) setup produced the lowest MSE and RMSE errors which indicates that this model has the highest performance compared to other models.


Keywords

COVID-19, Malaysia, ARIMA, LSTM, PROPHET.