Prediction of Infectious Diseases using Machine Learning: A Case Study of Tuberculosis in Nigeria

Abstract

Tuberculosis (TB) is one of the most deadly infectious diseases globally and remains a major public health challenge in Nigeria, which is classified among high TB-burden countries. TB transmission is strongly influenced by socio-economic conditions, population density, HIV co-infection, healthcare accessibility, and environmental factors. Early prediction of TB incidence is crucial for effective disease control, timely intervention, and optimal allocation of limited public health resources. In this study, different machine learning (ML) algorithms were trained using historical TB notification data, demographic indicators, socio-economic variables, healthcare access metrics, and environmental factors to predict TB trends in Nigeria. Four machine learning algorithms Random Forest, Support Vector Machine (SVM), Logistic Regression, and Neural Network models were developed and evaluated. Key features included population density, HIV prevalence, poverty index, healthcare facility coverage, historical TB incidence, temperature, and humidity. The results show that the Random Forest model outperformed other models, achieving superior predictive accuracy and sensitivity. This study demonstrates the effectiveness of machine learning as a powerful tool for tuberculosis prediction and surveillance in Nigeria.