Credit Risk Prediction Project

This repository contains a machine learning project focused on predicting loan default probabilities using a dataset from Kaggle. The goal of this project is to develop a credit risk model that helps identify high-risk borrowers, using techniques in data preprocessing, feature engineering, and predictive modeling.

Project Overview

Credit risk modeling is a crucial tool for lenders to evaluate the likelihood of borrowers defaulting on loans. In this project, I applied XGBoost, a high-performance gradient boosting algorithm, to build a predictive model that estimates the probability of loan default.

Key Steps

1. Data Preprocessing: Cleaned and prepared the data, handling missing values and outliers to ensure quality input for the model.
2. Feature Engineering: Selected and engineered relevant features to improve model performance and interpretability.
3. Model Training: Utilized XGBoost due to its efficiency and effectiveness in handling large datasets with complex patterns.
4. Evaluation: Assessed model performance using appropriate metrics to gauge predictive accuracy and stability.

Dataset

The dataset was sourced from Kaggle's Credit Risk Dataset, which includes borrower characteristics, credit history, and loan information. The target variable indicates whether a borrower defaulted on a loan, enabling us to frame the problem as a binary classification task.

Requirements

• Python 3.9.x
• Jupyter
• XGBoost
• Pandas
• Scikit-learn
• Matplotlib / Seaborn (for data visualization)

Install the dependencies via:

pip install -r requirements.txt

Usage

Open and run the notebook:

jupyter notebook main.ipynb

The notebook main.ipynb contains all steps of data exploration, model training, and evaluation, with detailed explanations and visualizations.

Results

The XGBoost model showed strong predictive capabilities, providing valuable insights into borrower risk levels. This model can serve as a basis for lenders to improve credit risk assessment and mitigate default risks.