The UTC Graduate School is pleased to announce that Gertrude Osei will present Master’s research titled, Effective Modeling of COVID-19 Outcomes Utilizing Google Trends Data: A VAR Approach on 03/02/2023 at 10:00am in Lupton Hall Room 302. Everyone is invited to attend.
Mathematics
Chair: Dr. Lani Gao
Co-Chair:
Abstract:
The World Health Organization (WHO) announced that COVID-19 was a pandemic disease as there were cases in several countries and territories. The United States is one of the hardest hit countries. Search engines, such as Google, provide useful real-time data from populations, and these data can be used as a near real-time indicator of public interest during a pandemic.Numerous researchers worked on forecasting the number of confirmed cases since anticipating the growth of the cases helps governments adopting knotty decisions to ease restrictions for their countries. This study forecasts the future patterns of daily new cases, cumulative cases, and deaths in the United States using Google Trends (GT) data on specific search terms connected to the COVID-19 pandemic and Center for Disease Prevention and Control (CDC) statistics on COVID-19’s spread. Search trends related to Long-Covid are also studied. A stationarity analysis of trend data was carried out with the use of an Augmented Dickey-Fuller (ADF) test in order to get rid of seasonality and trends. This analysis includes the computations of lag correlations between daily confirmed cases, cumulative death and Google trends data, Granger causality tests, and an out-of-sample forecasting exercise with competing Vector Auto Regression (VAR) models with a forecast horizon of eight periods ahead. To evaluate the performance of the proposed models, forecast accuracies were compared visually and using RMSE (root mean square error), MAE (Mean Absolute Error), MAPE (Mean Absolute Percentage Error), MASE (Mean Absolute Scaled Error).