🏥 Diabetes Prediction
Multi-Model ML Pipeline with Automated CI/CD
Models Trained
4
Dataset Size
442
Features
10
Best Model
XGBoost
📊 Model Performance Comparison
Comprehensive comparison of 4 machine learning models on diabetes prediction task
🤖 Models Compared
Linear Regression
Simple baseline model with linear assumptions
Ridge Regression
L2 regularization to prevent overfitting
Random Forest
Ensemble of 100 decision trees
XGBoost 🏆
Gradient boosting - typically the winner!
✨ Features
✅ Automated Training: All models trained automatically on every push
✅ Best Model Selection: System picks the best performer
✅ CI/CD Pipeline: Full automation with GitHub Actions
✅ Docker Ready: Containerized for easy deployment
✅ Flask API: REST API for predictions
✅ Comprehensive Tests: 7 automated tests ensure quality
🔄 CI/CD Pipeline
Every push to main branch automatically:
- Loads the diabetes dataset
- Trains all 4 models
- Runs comprehensive tests
- Generates performance visualizations
- Selects the best model
- Deploys updated dashboard
- Sends notifications (if enabled)
✅ Tests Passing
🚀 Auto-Deploy
📊 Live Updates
📚 Dataset Information
Diabetes Dataset (scikit-learn)
Ten baseline variables: age, sex, BMI, average blood pressure, and six blood serum measurements
Target: Quantitative measure of disease progression one year after baseline