Forecasting Covid 19 in Zimbabwe Using Time Series
- Author
- Mangoza, Praise
- Title
- Forecasting Covid 19 in Zimbabwe Using Time Series
- Abstract
-
Many people all over the world have been infected, and many others have died as a result of
the ongoing new Coronavirus outbreak around the world (Covid 19). In order to avert deaths,
it is critical to anticipate future infection cases and the viral transmission rate so that
healthcare providers can prepare ahead of time. The research community has an analytical
and difficult real-world difficulty in accurately projecting Covid 19 instances, hence the need
to employ various models to predict the dynamics of Covid 19 infection. In this study, the
researcher used daily level Covid 19 cumulative cases data for the entire country of
Zimbabwe to model the dynamics of Covid 19 infection. The data used was from March
2020 to January 31, 2022. ARIMA and LSTM forecasting models were used to model the
trends and make predictions. The mean absolute percentage error was used to assess the
models' effectiveness. Findings from this study have revealed that the LSTM model is more
effective at forecasting Covid 19 cases. The forecasting results have the potential to help
countries devise actions to stop the virus from spreading - Date
- 2022
- Publisher
- BUSE
- Keywords
- Forecasting
- Supervisor
- Miss. Pagan’a
- Item sets
- Department of Statistics and Mathematics
