Credit Card Approval ML Project

This repository hosts a Jupyter Notebook (credit-approval-ml-project.ipynb) that predicts credit card application approvals using machine learning.

Kaggle: https://www.kaggle.com/code/selvetelifdemirel/credit-approval-ml-project

📋 Table of Contents

• 🔍 Project Description
• 📂 Dataset
• ⚙️ Prerequisites
• 🚀 Installation
• ▶️ Usage
• 📝 Notebook Structure
• 📊 Results Summary
• 🤝 Contributing

🔍 Project Description

This project classifies credit card applications as "good" or "bad" based on applicant information and credit history. It covers:

• Data Preprocessing: Handling missing values and feature engineering (e.g., Age, WorkingYears).
• Imbalanced Data Handling: Balancing classes using SMOTE.
• Modeling: Training a baseline Random Forest classifier.
• Hyperparameter Optimization: Tuning model parameters with Optuna.
• Evaluation: Measuring performance using Accuracy, Precision, Recall, and F1-score.

📂 Dataset

The dataset consists of two CSV files, which should be merged on the applicant ID:

1. application_record.csv
   • Demographic and socio-economic features (e.g., CODE_GENDER, AMT_INCOME_TOTAL, NAME_EDUCATION_TYPE).
2. credit_record.csv
   • Monthly repayment status (MONTHS_BALANCE) and payment status codes.

⚙️ Prerequisites

• Python 3.7+
• Jupyter Notebook or JupyterLab
• Required Python libraries:
  • pandas
  • numpy
  • scikit-learn
  • imbalanced-learn
  • matplotlib
  • seaborn
  • optuna

🚀 Installation

1. Clone the repository:

   git clone https://github.com/<username>/credit-approval-ml-project.git

2. Install dependencies:

   pip install -r requirements.txt

▶️ Usage

1. Launch Jupyter Notebook:

   jupyter notebook

2. Open credit-approval-ml-project.ipynb.
3. Execute cells in order to follow data preprocessing, model training, and evaluation.

📝 Notebook Structure

1. Data Preprocessing
   • Load and merge CSV files
   • Analyze and clean missing values
   • Feature engineering (Age, WorkingYears)
2. Imbalanced Data Handling
   • Encode categorical variables
   • Apply SMOTE for oversampling
3. Modeling
   • Split data into training and testing sets
   • Train a baseline Random Forest model
   • Optimize hyperparameters with Optuna
4. Evaluation
   • Compare models and report performance metrics

📊 Results Summary

Below is the performance table for the Random Forest model optimized with Optuna:

	precision	recall	f1-score	support
0	0.97	0.98	0.98	85286
1	0.98	0.97	0.98	85110
accuracy			0.98	170396
macro avg	0.98	0.98	0.98	170396
weighted avg	0.98	0.98	0.98	170396

🤝 Contributing

1. Fork this repository.
2. Create a new branch:

   git checkout -b feature/your-feature

3. Commit your changes:

   git commit -m "Add some feature"

4. Push to your branch:

   git push origin feature/your-feature

5. Open a Pull Request.