Credit Scoring Model

Overview

This project was developed as part of the CodeAlpha Machine Learning Internship. The objective is to predict an individual's creditworthiness using historical financial data and machine learning classification algorithms.

Objectives

• Predict whether a customer is creditworthy.
• Analyze financial features affecting credit decisions.
• Compare multiple machine learning algorithms.
• Evaluate performance using standard classification metrics.

Dataset

The project uses a Credit Risk dataset containing customer financial information such as:

• Age
• Income
• Loan Amount
• Employment History
• Home Ownership
• Loan Intent
• Credit History Length
• Previous Defaults

Technologies Used

• Python
• Pandas
• NumPy
• Matplotlib
• Seaborn
• Scikit-Learn
• Google Colab

Machine Learning Models

1. Logistic Regression
2. Decision Tree Classifier
3. Random Forest Classifier

Results

Model	Accuracy
Logistic Regression	83.92%
Decision Tree	88.63%
Random Forest	92.98%

Random Forest Metrics

• Accuracy: 92.98%
• Precision: 97.52%
• Recall: 70.51%
• F1 Score: 81.84%
• ROC-AUC Score: 84.99%

Visualizations

• Confusion Matrix
• Feature Importance Analysis

Conclusion

Random Forest achieved the best performance with an accuracy of 92.98%, making it the most effective model for predicting customer creditworthiness in this project.

Author

Harshinee Shree G, Artificial Intelligence & Data Science Undergraduate