Skip to content
Stop guessing. Start growing.New: Digital Marketing Playbook 2026Increase your turnover today.Buy now →

Peer-reviewed · Explainable ML

Diabetes Risk Prediction Web App

A diabetes risk prediction web app that estimates risk and explains every prediction it makes. The work is published in a peer-reviewed journal (IJSRED, October 2022) and runs as an interactive Streamlit app.

  • Peer-reviewed (IJSRED, Oct 2022)
  • Deployed as a Streamlit web app
  • Live app: DiabetesRisk

By Adithya Reddy Nalla, Lead AI Engineer in Hyderabad.

Problem

A risk score without a reason is a number someone must take on faith. In healthcare that isn't enough: a clinician or patient needs to know why the model reached its answer.

For screening, the model also has to rank risk well across every threshold, not just score well at one.

Approach

  • Analysed the clinical dataset, then trained a Random Forest, which captures non-linear interactions between clinical features and needs no feature scaling.
  • Tuned hyperparameters with GridSearchCV using k-fold cross-validation, optimising ROC-AUC so the search rewards good risk ranking rather than raw accuracy.
  • Explained predictions with SHAP: a global view of which features drive risk across the population, and a local view of which values raised or lowered one person's score.
  • Deployed the model as an interactive Streamlit web app.
  • The live DiabetesRisk app offers three levels, from a no-labs lifestyle questionnaire to clinical values (glucose, HbA1c, blood pressure) and a comprehensive panel (lipids, liver, kidney). It shows each factor's contribution, updates results as you type, and includes an AI coach that turns the numbers into a personal plan. It is a screening and education tool, not a diagnosis.

Tech stack

Modelling
Pythonscikit-learnRandom ForestGridSearchCV
Explainability
SHAP
App
Streamlit

Results

  • 82%+

    Accuracy

  • 0.87

    ROC-AUC

  • 1

    SHAP explanation per prediction

Research software, not a clinical diagnostic tool.

Screenshots

  • DiabetesRisk web app on a phone: the heading 'Know your diabetes risk and what moves it', a 'Not a diagnosis' notice, three level tabs (Lifestyle, Clinical, Comprehensive) and a live score of 21 out of 100 with 'Glucose below prediabetes range'.
    The live DiabetesRisk app on a phone: three levels, a risk score that updates as you type, and a clear 'not a diagnosis' notice.

More projects

All projectsAbout Adithya Reddy Nalla