A Deep Learning Approach to Predict Malnutrition Status of 0-59 Month's Older Children in Bangladesh

Abstract

The state of malnutrition can be considered as a predominant issue for a developing nation like Bangladesh. Since today's children are the future's workforce, it explicitly impacts to the economic improvement of Bangladesh. So, prevention of child malnutrition is the most foremost investigation at this stage. The study aims to classify malnutrition based on deep learning approach of predictive modeling on significant malnutrition features to predict malnutrition status of a 0-59 months' older child. To do so an Artificial Neural Network (ANN) approach is applied to Bangladesh Demographic and Health Survey 2014 (BDHS) children data. This study clarifies how a predictive model classifies the malnutrition condition. ANN approach shows the best accuracy with wasting, underweight, and stunting. In conclusion, determining the malnutrition status using deep learning approach is the most scientific way to deal with it both for policymakers and clinicians.