How aditya-mankar/diabetes-prediction works
This is a diabetes-classification project with a model-development workflow and a Flask prediction interface. Training reads the diabetes data, preprocesses selected features, fits and evaluates a classifier, and saves a pickle; the Flask app loads that artifact and accepts patient inputs for prediction and a rendered result. The README also describes comparing algorithms, though the shown model source only demonstrates an SVC evaluation. The template and CSV contents were not sampled, so their details are not inferred here.
Main entry points: app.py Flask Routes and Prediction · model.py Preprocessing and Training · model.pkl Trained Model · index.html Prediction Form and Result
Components 9
Patient Data
- Diabetes Dataset
flask/diabetes.csv
Model Development
- Preprocessing and Training
flask/model.py - Model Evaluation
flask/model.py - Trained Model
flask/model.pkl
Prediction App
- Prediction Dataset
flask/diabetes.csv - Flask Routes and Prediction
flask/app.py
Web Presentation
- Prediction Form and Result
flask/templates/index.html - Page Styles
flask/static/css/style.css
Other components
- Patient / User
Connections 10
- Preprocessing and Training → Diabetes Dataset: reads data
flask/model.py - Preprocessing and Training → Model Evaluation: fits and evaluates
flask/model.py - Preprocessing and Training → Trained Model: saves model
flask/model.py - Flask Routes and Prediction → Trained Model: loads model
flask/app.py - Flask Routes and Prediction → Prediction Dataset: reads data
flask/app.py - Flask Routes and Prediction → Trained Model: predicts with
flask/app.py - Patient / User → Flask Routes and Prediction: submits inputs
flask/app.py - Flask Routes and Prediction → Prediction Form and Result: renders page
flask/app.py - Prediction Form and Result → Page Styles: uses styles
- Flask Routes and Prediction → Patient / User: returns result
flask/app.py