bandym05/AI-Enhanced-Health-Management-Combining-Predictive-Analytics-LLM-and-RAG-Driven-Support-for-Diabe

An intelligent system that combines predictive analytics, LLMs, and RAG to assess diabetes risk, offer lifestyle recommendations, and provide context-aware support using Eswatini health data. 🧠 Powered by explainable AI for smarter, localized health decisions. 🏆 Winner – IndabaX Eswatini 2024 Poster Presentation. # AI-Enhanced Health Management: Combining Predictive Analytics, LLM, and RAG-Driven Support for Diabetes and Heart Disease in Eswatini ## Demo Video is from a more improved and advanced version of this System: powered by Llamma 3.1 and Retrival Augmented Generation (RAG) Support github.com Welcome to the repository for the research project "Machine Learning-Driven Prediction and Management of Diabetes and Heart Disease" by Bandile Malaza, submitted for the Bachelor of Science in Information Technology at the University of Eswatini. # Authors - **Bandile Malaza** ## Project Overview #### This project focuses on leveraging machine learning algorithms to predict and manage diabetes and heart disease. The main objective is to develop a web application that uses predictive models to identify high-risk individuals and provide personalized recommendations for preventive care. The project aims to improve healthcare accessibility and empower healthcare providers to take a proactive approach to managing non-communicable diseases. #### The project includes the development of a user-friendly web application that allows users to input relevant health data for predicting the risk of diabetes and heart disease. The application provides personalized risk percentages, clear risk category labels, and lifestyle modification recommendations based on the predictions. ## Key Features - **Model Training and Evaluation:** The project involves training machine learning models for diabetes and heart disease prediction using datasets obtained from reputable sources like Kaggle and the UCI Machine Learning Repository. The models are evaluated based on metrics such as accuracy, precision, recall, F1 score, and AUC. - **Model Deployment:** After training the models using Google Colab, both the diabetes and heart disease models, along with the scalers, were saved to be used in the Flask web application for real-time …