Credit Risk Analysis Project

NBViewer link: https://nbviewer.org/github/cbayraktar24/Credit-Risk-Analysis/blob/main/Credit_Risk_Analysis_FINAL.ipynb

Project Overview

This project analyzes consumer loan data to identify key drivers of default risk and build a baseline credit risk model using logistic regression.

Dataset

• Source: https://www.kaggle.com/datasets/laotse/credit-risk-dataset
• Dataset file used locally: credit_risk_dataset.csv
• Observations: ~32,000 loans
• Target variable: loan_status (0 = non-default, 1 = default)

Tools Used

• Python (pandas, numpy)
• Data visualization (matplotlib, seaborn)
• Machine learning (scikit-learn)

Key Findings

• Default rates increase monotonically with higher interest rates.
• Higher income borrowers consistently show lower default risk.
• Default risk is highest for very large loans and also elevated for very small loans.
• Employment length shows a U-shaped relationship with default, with the lowest risk in mid-career borrowers.
• Loan percent of income is the strongest numeric predictor of default.

Modeling

• Built a baseline logistic regression model.
• Applied class weighting to address class imbalance and improve default detection.
• Demonstrated trade-offs between precision and recall based on business risk appetite.

How to Run

1. Clone this repository: git clone https://github.com/cbayraktar24/Credit-Risk-Analysis.git

2. Install required dependencies: pip install -r requirements.txt

3. Open the notebook: jupyter notebook Credit_Risk_Analysis_FINAL.ipynb