Credit Card Fraud Detection

University project for Big Data & Data Mining.

Project Description

End-to-end machine learning pipeline that detects fraudulent credit card transactions using the Kaggle Credit Card Fraud Detection dataset.

Models: Logistic Regression, Decision Tree Advanced technique: K-Means Clustering Imbalance handling: SMOTE

Setup

1. Clone or create the project folder

mkdir credit_card_fraud && cd credit_card_fraud
mkdir data src

2. Create virtual environment

python -m venv venv
source venv/bin/activate      # Mac/Linux
venv\Scripts\activate          # Windows

3. Install dependencies

pip install -r requirements.txt

4. Download the dataset

Download creditcard.csv from: https://www.kaggle.com/datasets/mlg-ulb/creditcardfraud

Place it in data/creditcard.csv.

Run

Terminal pipeline (prints summary)

python main.py

Streamlit dashboard

streamlit run app.py

Expected Output

• Dataset: 284,807 transactions, 492 fraud cases (~0.17%)
• Both models trained on SMOTE-balanced training data
• Evaluation on the original (imbalanced) test set
• Clustering summary showing fraud concentration per cluster
• Interactive Streamlit dashboard with all results