Please use this identifier to cite or link to this item: http://hdl.handle.net/123456789/3318
Full metadata record
DC FieldValueLanguage
dc.contributor.authorAlhassan, B.-
dc.date.accessioned2022-01-12T08:46:00Z-
dc.date.available2022-01-12T08:46:00Z-
dc.date.issued2021-
dc.identifier.urihttp://hdl.handle.net/123456789/3318-
dc.descriptionMASTER OF SCIENCE IN APPLIED STATISTICSen_US
dc.description.abstractThe Wa Regional Hospital’s live births and maternal mortality were analysed and forecasted in this study. The data were obtained from the Regional Hospital and covered the period from January 2009 to June 2020. This study applied Vector Autoregressive with exogenous variable to model the interdependent and dynamic structure that exists between the endogenous variables (live births and maternal mortality) and the exogenous variable (maternal age). Also, the dynamic structure of the observed data was explored using the Granger Causality. On the basis of the observed data, several competing models were examined, and the model selection criteria indicated that the VARX (13, 1) model was the most suitable model to represent the data generation process. The findings from the study indicated that the endogenous variables (live births and maternal mortality) do not Granger Cause each other. This is in confirmation that live births and maternal mortality cannot influence each other for better predictive accuracy or better still, neither variable (live births nor maternal mortality) can get better future values with the inclusion of past values of the other variable. The forecast results confirmed that the predicted values for both variables (live births and maternal mortality) mimics that of the observed values and with the predicted values falling within the acceptable limits.en_US
dc.language.isoenen_US
dc.titleMODELLING AND FORECASTING MATERNAL MORTALITY AND LIVE BIRTHS – USING VARX MODELS: CASE STUDY OF THE UPPER WEST REGIONAL HOSPITALen_US
dc.typeThesisen_US
Appears in Collections:Faculty of Mathematical Sciences



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