|
|
 |
| |
|
|
|
On Empirical Comparison of Forecast Performance of Autoregressive Moving Average Model and Generalized Autoregressive Conditional Heteroscedasticity Model |
|
|
|
PP: 41-49 |
|
|
|
doi:10.18576/jsapl/080105
|
|
|
|
Author(s) |
|
|
|
M. O. Akintunde,
A.S. Amusan,
A.O. Olawale,
|
|
|
|
Abstract |
|
|
| The federal Government of Nigeria in a bid to prevent infant mortality introduced Immunization programme whose aim is to prevent infectious diseases among the newly born children. However, the national programme on immunization (NPI) suffers recurrent setbacks due to many factors including ethnicity and religious beliefs. Nigeria is made up of 774 local Governments, 36 states with its federal capital in Abuja. The country is divided into six geo-political zones; north central, North West, North-East, South-East, South-West and South-South. The population is unevenly distributed across the country. The focus of this paper is to provide an understanding about the theoretical and practical application of ARMA and GARCH models to Nigeria immunization data as well as looking at the gains derivable from using either of the models. The paper compares the forecast performance of these two models and used performance measure indices to test the adequacy of the model that perform better. The data used was obtained from University College hospital, Ibadan annual reports on immunization. Augmented-Dickey Fuller test was used as a stationarity test for the series used, at level the series were not stationary but at first difference they were stationary, thereafter, the analysis of the data were performed. The results actually shown that the two models are good for modeling and forecasting the series under investigation, however, GARCH model slightly outperformed ARMA as shown by the analysis.
|
|
|
|
|
|
|
|
