Anomaly Detection for Credit Card Fraud: A Deep Learning Approach

Project Summary

This repository showcases the project for the EL-GY-9163 Machine Learning for Cyber-security course at New York University. The focus is on replicating and improving a deep learning model for detecting fraud in credit card transactions, inspired by the research of Tung-Yu Wu and You-Ting Wang.

Execution Instructions

• Dataset Acquisition:

  • Download the Credit Card Fraud Detection dataset from Kaggle.
  • Save the dataset in a directory within the project folder.

• Notebook Configuration:

  • Update the dataset path in main.ipynb to reflect the location of your downloaded dataset.

• Notebook Execution:

  • Run main.ipynb in Jupyter Notebook to review the project's implementation and results.

Acknowledgments

We thank Prof. Marco Romanelli, Prof Alexandre Araujo, Prof Naman Patel, and Prof Chinmay Nerurkar and the teaching assistants for their guidance, mentorship, and expertise throughout this project. Also, credit to the original research available on GitHub.