How vaibhavgovindwar/diabetes-prediction-app works
This repository is a Streamlit diabetes-risk prediction app, with a notebook and Pima dataset representing its model-development context. The main user flow is to enter eight measurements, request a prediction, and see a binary label and probability. At runtime, the app downloads model and scaler files from Google Drive when absent, scales the inputs, and calls the model. The app source is excerpted; training details and the background image are not, so those parts are limited to README-supported relationships.
Main entry points: app.py Streamlit app · Pima indians diabetes dataset.csv Pima dataset · pima-indian-diabetes-binary-classification.ipynb Training notebook
Components 8
Model development
- Pima dataset
Pima indians diabetes dataset.csv - Training notebook
pima-indian-diabetes-binary-classification.ipynb
Prediction runtime
- Google Drive
- Feature scaler
app.py - Keras model
app.py
User interface
- Streamlit app
app.py - Prediction result
app.py
Other components
- User
Connections 6
- Pima dataset → Training notebook: trains on
README.md - User → Streamlit app: enters measurements
app.py - Streamlit app → Google Drive: downloads artifacts
app.py - Streamlit app → Feature scaler: scales inputs
app.py - Streamlit app → Keras model: predicts with
app.py - Streamlit app → Prediction result: displays
app.py