Time Series Models for the Decrease In Under-Five Mortality Rate in Ghana Case Study 1961 - 2012

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
A time series data, comprising of annual estimates of Under-five Mortality rates for Ghana from the year 1961 to 2012, obtained from the Worldbank website is used for the analysis. Three time series models; the Box-Jenkins (ARIMA), the Bayesian Dynamic Linear Model, and the Random walk with drift models are built for the decline of Ghana‟s under-five Mortality. Each model is built with data values from 1961 to year 2000, and an in-sample forecasting is made with each model from year 2001 to 2012. The Mean Squared Error (MSE) and the Mean Absolute Percentage Error (MAPE) as a measure of accuracy are used to determine the best fit model. The Random Walk with drift model produced the least values for both the MSE and the MAPE and is selected the best fit Model, and used for an out-of-sample forecasting for the years (2013 – 2016), producing respectively; 69.3, 66.6, 64.0 and 61.3 deaths per 1,000 live births. The forecast value of 64.0 deaths per 1000 live births for year 2015 shows that Ghana may not be able to realize her Millennium Development Goal four (MDG 4) target of reducing her Under-five Mortality rate to about 42.7 deaths per 1,000 live births by the year 2015.
A Thesis submitted to the Department of Mathematics, Kwame Nkrumah University of Science and Technology in partial fulfillment of the requirements for the degree of Master of Science in Industrial Mathematics, 2014