Associating Perinatal Mortality With Diet By Adapting Robust Clustering Using Links For Categorical Variables

Abstract

Perinatal Mortality (perinatal death), is death of a neonate within 6 days (early neonatal mortality) or from 7 – 27 days of birth (late neonatal mortality). Food consumed by an expectant mother is said to have an impact on the pregnancy outcome apart from other factors. For the past few years, perinatal mortality rate has been increasing in developing and under-developed parts of the world. Two-thirds of the world’s perinatal deaths occur in only 10 countries, and Pakistan is ranked third amongst these countries. These deaths have not been studied widely, in fact they have been under-reported and these reports have not even been considered in any attempts made to improve birth outcomes in developing nations [1]. Nutritional, socioeconomic, demographic and health advice seeking behavior factors are responsible for higher mortality rates in countries such as Pakistan. Data mining and machine learning can be used to identify factors that are responsible for such high infant mortality rates as it is an important factor indicating progress on Millennium Development Goals. In this paper, we discuss how using ROCK we can cluster expectant mothers as per the food intake and identify major food items causing perinatal mortality.