This project automates bank credit risk assessment using AI and machine learning models to predict loan defaults. It streamlines the credit process with predictive analytics, model evaluation, explainability (SHAP), and deployment readiness.
๐ค Loan Default Prediction โ Automating Bank Credit Processes Using AI
This project applies AI and machine learning to predict whether a customer is likely to default on a loan, enabling banks to make informed and automated credit decisions.
๐ Problem Statement
Banks often struggle to accurately assess credit risk and loan default probability. This project automates that process by using classification models to predict default based on applicant data.
๐ง Key Features
- Predicts loan defaults using historical banking data
- Compares multiple ML algorithms (Logistic Regression, Random Forest, XGBoost, etc.)
- Evaluates models via ROC-AUC, F1 score, Precision, and Recall
- Provides explainability with SHAP (SHapley Additive exPlanations)
- Visualization of feature importance and model decisions
- End-to-end pipeline from data preprocessing to model interpretation
๐งฐ Tech Stack
- Python
- scikit-learn
- XGBoost
- SHAP
- pandas, matplotlib, seaborn
๐ ๏ธ Setup Instructions
- Clone the Repository
git clone https://github.com/aimaster-dev/defaultloanprediction.git
cd defaultloanprediction
- Install Dependencies
<pre><code class="lang-bash">pip install -r requirements.txt</code></pre>
- Run the Notebook
Open loandefaultprediction.ipynb in Jupyter or run it using:
<pre><code class="lang-bash">jupyter notebook loandefaultprediction.ipynb</code></pre>
๐ Model Evaluation
- Best Model: XGBoost
- ROC-AUC: High discriminative power
- SHAP Values: Used to interpret individual predictions and global feature impact
๐ Highlights
- Automated model selection and tuning
- Transparent credit risk scoring using SHAP
- Business-focused evaluation for banking applications
๐ Resources
- ๐ Project Report (PDF)
๐ License
MIT License. See
LICENSE` for details.
๐ Acknowledgments
Thanks to aimaster-dev for sharing this impactful project in AI-driven finance.
