Please use this identifier to cite or link to this item: http://hdl.handle.net/123456789/1167
Title: COMPARATIVE ANALYSIS OF SARIMA AND SETAR MODELS IN PREDICTING PNEUMONIA CASES IN THE NORTHERN REGION OF GHANA
Authors: Badawi, M.
Issue Date: 2016
Abstract: Acute respiratory infection (ARI) range, in spectrum, from mild colds and coughs to life-threatening pneumonias. ARI particularly pneumonia is the major cause of morbidity and mortality among young children under five in developing countries with Ghana not an exception. In this study, we compared the linear SARIMA Model and the nonlinear two regime SETAR Model in predicting pneumonia cases in northern region of Ghana. Data on monthly pneumonia cases obtained from the Tamale Teaching Hospital database was modeled using SARIMA and SETAR models. The results revealed that SARIMA (1, 1, 1)(0, 0, 1)12 model was the best SARIMA model for the pneumonia cases. This model has the least AIC of 83.50, AICc of 83.71 and BIC of 96.08. Also, SETAR (2; 4, 3) model was identified as the best SETAR model. This model has an AIC of -165.42 and BIC of 90.06 as the least among the possible models based on the grid search for the best model. Diagnostic checks of both models with the Ljung-Box test and ARCH-LM test revealed that both models were free from higher-order serial correlation and conditional heteroscedasticity respectively. Based on the forecast assessment from the linear SARIMA and the nonlinear SETAR models, the forecast measures suggest that the nonlinear SETAR model outperform the linear SARIMA model. A comparative analysis of the forecasting accuracy of these models with the Diebold-Mariano test revealed that there were no significant difference in the forecasting performance of the two models. The two models were therefore proposed for predicting Pneumonia cases in the region.
URI: http://hdl.handle.net/123456789/1167
Appears in Collections:Faculty of Mathematical Sciences

Files in This Item:
File Description SizeFormat 
COMPARATIVE ANALYSIS OF SARIMA AND SETAR MODELS IN PREDICTING PNEUMONIA CASES.pdfMASTER OF SCIENCE DEGREE IN APPLIED STATISTICS1.01 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.