Building Predictive Models for Infant Mortality Rate in Bangladesh

Infant Mortality Rate (IMR) is one of the key indicators for a developing country, Bangladesh in particular, to measure the health and social status of its population. Many risk factors are associated with IMR that requires close investigation. This paper aims to build predictive models by identifying such potential risk factors responsible for infant mortality in Bangladesh. The World Bank data repository, based on "World Development Indicator (WDI) 2015", has been analyzed. We identify significant factors, build multiple stepwise linear regression models, and evaluate crosscoefficient model performances using MAE, MRE and MSE under the shed of k-fold cross validation technique. The performance of the final prediction model yields significant outcome towards the policy makers that could eventually reduce the infant mortality rate in Bangladesh.