Glucobuddy: Detecting Diabetes Risk Using Machine Learning

Diabetes mellitus is a chronic disease affecting over 420 million people globally, contributing significantly to mortality, disability, and healthcare costs. Early detection and risk assessment are critical in preventing severe complications such as cardiovascular disease, kidney failure, and neuropathy. Traditional diagnostic approaches, including fasting glucose and HbA1c testing, require medical infrastructure and trained personnel, making them difficult to access in resource-limited areas. This thesis presents Glucobuddy, an intelligent system designed to predict diabetes risk levels using machine learning models and to enhance user interaction through an integrated AI chatbot. The system analyzes key health indicators including age, glucose levels, and body mass index (BMI) to classify individuals into low-risk or high-risk categories. Three machine learning algorithms—Logistic Regression, Random Forest, and Support Vector Machines (SVM)—are evaluated and compared using performance metrics such as accuracy, precision, recall, and F1-score. In addition to automated risk classification, Glucobuddy incorporates an AI-powered chatbot designed to communicate results, provide general diabetes education, answer common queries, and suggest preventive actions based on the user’s risk profile. This interactive approach aims to enhance user understanding and engagement. The proposed system offers a cost-effective, accessible, and scalable solution for early diabetes risk screening, with particular focus on underserved communities. It provides healthcare professionals and individuals with a practical tool for early intervention, contributing to improved health outcomes and reduced healthcare burdens.