colinmarklubembe/tyhpoid-analysis-in-uganda

# Typhoid Analysis in Uganda with a Comparative Study of Machine Learning Algorithms ## Abstract Typhoid continues to be a public health issue in Uganda with over 5,000 incidences every year. Because of the inadequate health facilities, the rise of such a disease is quite alarming. It is essential for us to have proper analysis and prediction of the patterns and occurrences of Typhoid. This report presents a comparative analysis of different machine learning models for predicting Typhoid percentages among the population in the different districts of Uganda. For this particular regression problem, we evaluated the performance of five algorithms that is to say Linear Regression Model, Support Vector Regression model, Decision Tree, Random Forests and the Neural Networks Regression Model. Using a data set of Typhoid Incidences, Environmental factors and population our results show that the Neural Networks is the most optimal model. Each model was evaluated with the Mean squared error, the Root Mean squared error, the mean absolute error and the R2 score. ## Keywords Typhoid, Machine Learning, Regression, Linear Regression Model, SVR, Decision Tree, Random Forest, Neural Networks. ## Introduction Typhoid fever, a waterborne disease caused by Salmonella Typhi, remains a significant public health concern in Uganda, with over 50,000 citizens attacked annually. The disease is often associated with poor sanitation, contaminated water, and inadequate hygiene. In Uganda, the disease poses a significant burden on the health care system, particularly in urban areas like Kampala, where population density and poor living conditions contribute to the fast spread of the disease. This study aims to analyze and predict the percentage of people with typhoid in the different districts in Uganda using various regression models including the Linear Regression, Decision Tree, Random Forest, Support Vector Regression (SVR), and Neural Networks. The models' performance is evaluated using …