How akashavcoewala/churn-prediction-using-nlp-ml works
This repository is a telecom churn-prediction project: a Flask web app accepts customer details and returns a churn result, while standalone scripts train and evaluate models from the dataset. The app also generates a PDF report. The README documents the saved XGBoost model and web prediction capability. Some app code is omitted, so model-loading and template-routing details are represented only where documented; the report-generation call is visible in the excerpt.
Main entry points: app.py Flask routes · churn2.csv Churn dataset · XGB.py XGBoost training · xgboost_model.joblib Saved XGBoost model
Components 14
Prediction web app
- Flask routes
app.py - Login and registration
app.py - Customer form
app.py - Feature preparation
app.py - Churn prediction
app.py - Prediction result
app.py - PDF report generation
app.py
Model development
- XGBoost training
XGB.py - XGBoost evaluation
XGB2.py - Random Forest baseline
RandomForest.py
Data and artifacts
- Churn dataset
churn2.csv - Saved XGBoost model
xgboost_model.joblib - Churn report PDF
Churn_report.pdf
Other components
- Customer
Connections 13
- Customer → Flask routes: uses
README.md - Flask routes → Login and registration: routes to
app.py - Flask routes → Customer form: routes to
app.py - Customer form → Feature preparation: provides fields
app.py - Feature preparation → Churn prediction: prepares features
app.py - Churn prediction → Saved XGBoost model: uses model
README.md - Flask routes → Prediction result: renders result
app.py - Flask routes → PDF report generation: generates report
app.py - PDF report generation → Churn report PDF: writes PDF
app.py - Churn dataset → XGBoost training: training input
XGB.py - Churn dataset → XGBoost evaluation: training input
XGB2.py - Churn dataset → Random Forest baseline: training input
RandomForest.py - XGBoost training → Saved XGBoost model: produces model