Abstract:
This study focuses on developing a predictive machine learning model to detect non-communicable diseases (NCDs) in low-resource settings: case study Kitui county, Kenya. Poor access to diagnostics in these settings usually results into delayed diagnosis and unfavorable health outcomes. The study combines the data of healthcare to determine the main risk factors and test the machine learning algorithms, such as logistic regression, random forests, and gradient boosting, based on their accuracy and clinical significance. The research design employed retrospective type of research, supervised learning methods are employed in data preprocessing, feature selection, model training and validation, and enhanced performance in terms of accuracy, precision, recall, and F1-score. An intelligent hybrid machine learning model was developed with the accuracy of 0.93. It is meant to enhance early diagnosis, resource utilization and healthcare expenses.
