How bashutosh017/credit-card-approval-ml-model works
This repository is a Flask web application for estimating credit-card approval likelihood. The main workflow is an applicant entering financial details in a two-step browser form, submitting them to the prediction endpoint, and receiving an approval estimate. The server validates and normalizes inputs, applies the serialized LightGBM model, and returns a result. It also serves landing and documentation pages. The README describes separate training and console-prediction modules, but those files are absent from the supplied tree; this map therefore covers the implemented web runtime only.
Main entry points: server.py Flask routes · app.js Assessment form · index.html Landing page · docs.html Documentation page
Components 14
Web interface
- Landing page
templates/index.html - Documentation page
templates/docs.html - Assessment page
templates/check.html - Assessment form
static/js/app.js - Client validation
static/js/app.js - Result display
static/js/app.js
Prediction service
- Flask routes
server.py - Prediction endpoint
server.py - Input validation
server.py - Feature preparation
server.py - Model inference
server.py
Inference assets
- LightGBM model
Lightgbm_credit2.pkl - Model features
Model_feature2.pkl
Other components
- Applicant
Connections 14
- Applicant → Flask routes: requests pages
server.py - Flask routes → Landing page: renders
server.py - Flask routes → Documentation page: renders
server.py - Flask routes → Assessment page: renders
server.py - Applicant → Assessment form: enters details
static/js/app.js - Assessment form → Client validation: validates fields
static/js/app.js - Client validation → Prediction endpoint: posts JSON
static/js/app.js - Prediction endpoint → Input validation: validates payload
server.py - Input validation → Feature preparation: normalizes fields
server.py - Feature preparation → Model inference: prepares matrix
server.py - Model inference → LightGBM model: predicts probabilities
server.py - Model inference → Model features: orders features
server.py - Prediction endpoint → Result display: returns JSON
static/js/app.js - Result display → Applicant: shows estimate
static/js/app.js