💳 Credit Score Classification System

A machine learning-powered web application that predicts customer credit scores using advanced classification algorithms. Built with Python, Streamlit, and Random Forest.

---

📋 Table of Contents

• Overview
• Features
• Demo
• Installation
• Usage
• Project Structure
• Model Information
• Tech Stack
• Contributing
• License

---

🎯 Overview

This project implements a complete credit score classification pipeline that:

• Analyzes customer financial data to predict credit worthiness
• Classifies customers into three categories: Good, Standard, and Poor
• Provides an interactive web interface for real-time predictions
• Offers actionable recommendations based on predictions

---

✨ Features

• 🤖 Machine Learning Model: Random Forest classifier trained on comprehensive credit data
• 🌐 Interactive Web App: Modern, responsive Streamlit interface with dark theme
• 📊 Real-time Predictions: Instant credit score classification with probability scores
• 💡 Smart Recommendations: Tailored suggestions based on prediction outcomes
• 📈 Visual Analytics: Interactive probability charts using Plotly

---

🤗 Hugging Face Resources

Resource	Link
Model Card	AdityaaXD/credit-score-classifier
Dataset Card	AdityaaXD/Credit-Score-Classification

---

🖥️ Demo

Main Prediction Interface

The application features a premium dark-themed UI with:

• Easy-to-use input forms for customer data
• Real-time prediction results with confidence scores
• Probability distribution visualization

---

🚀 Installation

Prerequisites

• Python 3.8 or higher
• pip package manager

Setup Steps

1. Clone the repository

   git clone https://github.com/yourusername/credit-score-classification.git
   cd credit-score-classification

2. Create a virtual environment (optional but recommended)

   python -m venv venv
   source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies

   pip install -r requirements.txt

4. Run the application

   streamlit run app.py

5. Open your browser and navigate to http://localhost:8501

---

📖 Usage

Making Predictions

1. Navigate to the Predict tab in the application
2. Enter customer information:
   • Numerical Features: Age, Annual Income, Number of Bank Accounts, etc.
   • Categorical Features: Occupation, Credit Mix, Payment Behavior, etc.
3. Click "Predict Credit Score"
4. Review results: View the prediction, confidence scores, and recommendations

Input Features

Feature	Type	Description
Age	Numerical	Customer's age
Annual_Income	Numerical	Yearly income in currency
Monthly_Inhand_Salary	Numerical	Monthly take-home salary
Num_Bank_Accounts	Numerical	Total bank accounts owned
Num_Credit_Card	Numerical	Number of credit cards
Interest_Rate	Numerical	Average interest rate on credit
Num_of_Loan	Numerical	Number of active loans
Delay_from_due_date	Numerical	Average delay in payments (days)
Outstanding_Debt	Numerical	Total outstanding debt
Credit_Utilization_Ratio	Numerical	Credit utilization percentage
Credit_History_Age_Months	Numerical	Credit history length in months
Occupation	Categorical	Employment type
Credit_Mix	Categorical	Types of credit accounts
Payment_Behaviour	Categorical	Spending patterns
Payment_of_Min_Amount	Categorical	Minimum payment behavior

---

📁 Project Structure

Credit Score Classification/
├── 📄 app.py                    # Streamlit web application
├── 📓 Credit_Score_Classification.ipynb  # Jupyter notebook with EDA & training
├── 📄 requirements.txt          # Python dependencies
├── 📄 README.md                 # This file
├── 📊 train.csv                 # Training dataset
├── 📊 test.csv                  # Test dataset
└── 📊 submission.csv            # Model predictions

🤗 Model Artifacts (Hosted on Hugging Face)

The trained model artifacts are hosted on Hugging Face Hub and are automatically downloaded when you run the application:

File	Description
models/final_model.pkl	Random Forest classifier
models/scaler.pkl	Feature scaler
models/label_encoder.pkl	Target label encoder
models/feature_info.pkl	Feature metadata
models/onehot_encoder.pkl	Categorical encoder

---

🧠 Model Information

Algorithm

• Model: Random Forest Classifier (n_estimators=100)
• Target Classes: Good, Standard, Poor

Feature Engineering

• Numerical Features: 17 scaled features including financial metrics
• Categorical Features: 5 one-hot encoded features
• Preprocessing: StandardScaler for numerical, OneHotEncoder for categorical

Performance

The model was trained on a comprehensive credit dataset with features covering:

• Payment history
• Credit utilization
• Account composition
• Financial behavior patterns

---

🛠️ Tech Stack

Category	Technology
Language	Python 3.8+
ML Framework	Scikit-learn
Web Framework	Streamlit
Model Hosting	Hugging Face Hub
Data Processing	Pandas, NumPy
Visualization	Plotly, Matplotlib, Seaborn
Development	Jupyter Notebook

---

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

---