Credit Card Fraud Detection using Machine Learning

Capstone Project for CS 6140

---

Directory Structure

• Report/
  Contains the final report for this project.
• Deliverables/
  Contains all the final Jupyter Notebooks for our exploratory data analysis and all the models.
• Presentation/
  Contains the presentation decks used for proposal and final presentations.
• Trials and experimentation/
  Contains Jupyter Notebooks that we wrote over the course of working on this project. These contain experiments, trials and their results, some of which failed. The final notebooks were built atop the ideas and approaches that were explored via these rough notebooks.

Packages Required

pandas
numpy
matplotlib
scikit-learn
pytorch
imbalanced-learn
feature-engine
seaborn
plotly

Kaggle API Key

The dataset is directly downloaded from Kaggle using the Kaggle CLI. This requires an API key to be generated and used in all of our notebooks. We have added placeholder values for Kaggle username and API keys in the first cell of our notebooks which must be edited before running.