Diabetes Risk Screener

• 

  Accuracy Results

• 

  Diabetes 130 US hospitals for years 1999-2008 Dataset Snippet

• 

  Website1

• 

  Website2

• 

  Model training code snippet

Inspiration

We wanted to make preventive healthcare more accessible. Many people don’t realize they may be at risk for diabetes until it’s too late, so we built a quick and easy tool to raise awareness.

What it does

The app predicts a user’s risk of prediabetes or diabetes in seconds. By entering basic lifestyle and health data like age, BMI, and activity level, users instantly get a risk prediction powered by a machine learning model.

How we built it

We used Flask for the backend, scikit-learn for training an SGD-SVM model, and Joblib for model storage. The frontend was made with HTML, CSS, and JavaScript to provide a clean, responsive form interface.

Challenges we ran into

Matching the model’s feature order, scaling input data correctly, and managing form validation were tricky. We also had to ensure the app stayed accurate while remaining simple for users.

Accomplishments that we're proud of

We built a working ML-powered web app that gives real-time health insights. Additionally, good precision, recall, f1-score, and confusion matrix results for the kind of task.

What we learned

We learned how to integrate a trained ML model into a Flask app, handle real-world input validation, and design user-friendly health tools responsibly.

What's next for Diabetes Risk Screener

We plan to improve model accuracy, add visual analytics for user insights, and expand to detect other health risks like heart disease or hypertension.

Built With

• flask
• github
• google-colab
• kaggle
• python
• scikit-learn