Monitoring the COVID-19 Spread via Regression Models with ARIMA Errors  
Author Huifen Chen




Abstract Statistical methods have been used on outbreak detection in the public health. In the past three years more than 670 million people were infected with COVID-19 and more than 6.8 million people were dead worldwide. Despite of the lower fatality rate of the major circulating omicron variant now, surveillance tool for outbreak detection is needed for preventing future possible pandemic due to new variants. ARIMA models are popular time-series methods for forecasting the spread of infectious diseases such as COVID-19. However, the spread also depends on other factors. We propose a regression model with an ARIMA error term for forecasting the spread of COVID-19 and use Taiwan data for illustration. Our results show that all the daily numbers of cases fall within the prediction intervals, which indicates no outbreak signal for that time period.


Keywords ARIMA, COVID-19, prediction, outbreak, surveillance
    Article #:  RQD28-242

Proceedings of 28th ISSAT International Conference on Reliability & Quality in Design
August 3-5, 2023