Diabetes is one of the most rapidly growing chronic diseases worldwide and poses a significant threat to global health. Early diagnosis plays a crucial role in preventing severe complications such as cardiovascular diseases, kidney failure, and nerve damage. However, traditional diagnostic methods often rely on clinical tests and expert consultation, which can be time-consuming and sometimes inaccessible to all individuals.

This paper presents a machine learning-based diabetes prediction system designed to provide early and accurate prediction using patient health data. The system utilizes key medical attributes such as glucose level, blood pressure, body mass index (BMI), insulin level, age, and other relevant features. Advanced machine learning algorithms, namely Random Forest and XGBoost, are employed to improve prediction accuracy and handle missing values effectively.

The proposed system also includes a user-friendly web interface developed using Streamlit, allowing users to input their health parameters and obtain instant predictions. The results demonstrate that the system achieves better accuracy compared to traditional models, making it a reliable tool for early detection and preventive healthcare.