Trend Analysis and Determinants of under-5 Mortality in Nigeria: A Machine Learning Approach

The study aimed to examine the trend of under-five mortality rate in Nigeria from 2003 2018 and determinants using Demographic Health Survey (NDHS) data. data for was conducted 2003, 2008, 2013, 2018. These four surveys were used trends within period, while machine learning applied only dataset being latest Nigeria. partitioned into training testing sets. 30% randomly selected testing, 70% model. Before applying logistic regression neural networks, essential variables first a random forest classifier.
 showed that rates 200.72, 156.86, 128.05, 132.02 respectively, per 1,000 live births. This result means one every five children died before their fifth birthday six eight seven forecast indicated would likely be 102.17 2023. variable importance breastfeeding (when child put breast after birth) had highest contribution mortality. breakdown delaying 6-23 hours comparison with 0-5 birth increases by 1.4 fold likelihood death. accuracy (LR) on test set 60%, deep network (DNN) 74%, recall (sensitivity) LR 63%, DNN 75%), Precision (LR=97%, DNN=95), F1 score (LR=76%, DNN=84%) area under curve (AUC) (LR=79%, DNN=77%).
 Both models performed very well discriminative ability accuracy. better performance than regression.