Credit Risk Explainability: Explainable AI for Loan Approval

Overview

This repository provides an open-source implementation of the research paper:
"An Explainable AI Framework for Credit Evaluation and Analysis" (Applied Soft Computing Journal, 2024).
📄 Paper Link: Elsevier - Applied Soft Computing

The project applies Machine Learning (ML) and Explainable AI (XAI) techniques to improve transparency in credit risk assessment and loan approval decisions. Using Random Forest, SHAP, LIME, and Partial Dependency Plots (PDP), it explains why a loan is approved or rejected, helping financial institutions and borrowers understand decision-making processes.

Features

• Implements Logistic Regression, Decision Tree, and Random Forest for credit risk evaluation.
  
• Uses LIME and SHAP to provide both local and global explanations of model decisions.
  
• Partial Dependency Plots (PDP) to visualize feature influence on predictions.
  
• Random Forest achieves 99.8% accuracy in loan approval classification.
  
• Open-source implementation with structured, modular, and well-documented code.
  

Repository Structure

Credit Risk Explainability/ │── Credit Risk Explainability.ipynb # Google Colab notebook with full implementation │── requirements.txt # Dependencies list │── README.md # Documentation │── LICENSE # Open-source license │── CITATION.md # Paper citation details

Getting Started

Clone the Repository

bash git clone https://github.com/your-username/Credit-Risk-Explainability.git cd Credit-Risk-Explainability

Install Dependencies

bash pip install -r requirements.txt

Open in Google Colab

To run the notebook interactively, use the link below:

Open in Google Colab

Results

| Model | Accuracy | Precision | Recall | F1-Score | |-----------------|---------|----------|--------|---------| | Logistic Regression | 0.980 | 0.980 | 0.980 | 0.980 | | Decision Tree | 0.997 | 0.996 | 0.995 | 0.997 | | Random Forest | 0.998 | 0.997 | 0.997 | 0.998 |

Explainability Results

• SHAP Explainer: Provides both local and global explanations for feature importance.
  
• LIME Explainer: Highlights feature contributions for individual predictions.
  
• PDP Plot: Demonstrates the relationship between input features and loan approval decisions.
  

Citation

If you use this repository in your work, please cite the original paper:
bibtex @article{Nallakaruppan2024XAI, title={An Explainable AI framework for credit evaluation and analysis}, author={M.K. Nallakaruppan, Balamurugan Balusamy, et al.}, journal={Applied Soft Computing}, year={2024}, doi={10.1016/j.asoc.2024.111307} }

License

This project is licensed under the MIT License. See the LICENSE file for details.

Contributing

Contributions are welcome. If you find issues or improvements, feel free to open an issue or submit a pull request.