Credit Card Risk Classification for Applicant Evaluation

A web-based application that predicts credit card approval risk using machine learning model.

Class of RSBP C

Name	NRP
Nathaniel Ryo Kurniadi	5025221019
Mikha Gracia Sugiono	5025221037
Moch. Avin	5025221061
Azarel Grahandito Adi	5025221126
Muhammad Nabil Fadhil	5025221200

About this Project

This application is designed to help financial institutions assess credit card applications by predicting the risk level of potential customers. Built using Flask and CatBoost machine learning model, it provides real-time risk assessment based on various applicant parameters.

Key Features:

• Interactive web interface for data input
• Real-time risk prediction using CatBoost model
• Comprehensive applicant assessment
• Responsive design for all devices
• Docker support for easy deployment
• Secure and efficient data processing

Technologies Used:

• Python/Flask for backend
• CatBoost ML Model for predictions
• HTML/CSS/JavaScript for frontend
• Docker for containerization
• Pandas/NumPy for data processing

(Back to Top)

How it works

Input Parameters

The application processes the following input parameters:

1. Personal Information

   • Gender (Male/Female)
   • Car Ownership Status
   • Real Estate Ownership
   • Number of Children

2. Contact & Education

   • Education Level (4 categories)
   • Phone Ownership (Work/Personal)
   • Email Status

3. Financial & Employment

   • Income Level (3 ranges)
   • Years of Employment
   • Employment Status
   • Occupation Type
   • Family Member Count

Risk Assessment Process

1. Data Collection & Validation

   • Web form captures user inputs
   • Input validation ensures data quality
   • Data preprocessing for model compatibility

2. Model Processing

   • CatBoost model evaluates risk factors
   • Real-time prediction generation
   • Binary classification (High Risk/Low Risk)

Prerequisites

Before you begin, ensure you have:

• Python 3 or higher installed
• Docker (if using containerized deployment)
• A modern web browser
• Active internet connection

Installation Guide

Option 1: Using Docker (Recommended)

1. Clone the repository:

git clone https://github.com/NathanielRN/CreditCardApprovalPrediction.git
cd CreditCardApprovalPrediction

2. Build and run with Docker:

docker build -t credit-risk-app .
docker run -p 5000:5000 credit-risk-app

3. Access the application at http://localhost:5000

Option 2: Manual Setup

1. Clone the repository:

git clone https://github.com/NathanielRN/CreditCardApprovalPrediction.git

cd CreditCardApprovalPrediction

2. Create and activate a virtual environment:

# Create virtual environment
python -m venv venv

# Activate virtual environment

# For Windows:
venv\Scripts\activate
# For Unix/MacOS:
source venv/bin/activate

3. Install dependencies:

pip install -r requirements.txt

4. Run the application:

flask run

5. Access the application at http://localhost:5000