Machine Learning-Based Diabetes Risk Prediction Using Random Forest and Streamlit Interface

This project focuses on developing an efficient, and user-friendly diabetes mellitus checkup application powered by machine learning. The application is designed to offer a simple and effective means through which people can determine risk of diabetes, a condition that continuous to impact millions globally. By analyzes key health parameters such as glucose levels, BMI, insulin level, and age to predict diabetes risk using a Random Forest Classifier. A seamless and intuitive interface is created using Streamlit, a modern web application framework, allowing users to input their health data effortlessly. A user-friendly interface built with Streamlit allows seamless interaction, enabling patients to input health data and receive instant feedback, including graphical visualizations comparing their metrics to population data. The interface delivers instant feedback, including personalized diabetes risk predictions and graphical visualization that compare user metrics to population-level benchmarks. This feature enhances user understanding and engagement by providing actionable insights. The system architecture integrates advanced data analysis techniques with real-time processing capabilities. The prototype is designed using Python-based machine learning models and incorporates a smartphone-compatible interface for data storage and visualization. The application has been tested with case scenarios to ensure reliability and usability. This paper presents a robust, interactive system for early diabetes detection, with potential for deployment in various healthcare settings to promote personalized health monitoring and timely intervention. The project demonstrates the transformative potential of integrating machine learning with user-centred design in addressing pressing healthcare challenges.