Credit Card Fraud Detection Dashboard

This project creates a Streamlit-powered interactive dashboard for analyzing credit card fraud patterns through network visualization and transaction analysis.

View Dashboard Demo

Dataset Information

This dashboard uses the Credit Card Fraud Detection Dataset from Kaggle:

• Contains transactions made by European cardholders in September 2013
• 284,807 transactions with 492 frauds (highly imbalanced dataset)
• Features V1-V28 are PCA-transformed for confidentiality
• Only features not transformed are 'Time' and 'Amount'
• 'Class' is the target variable (1 = fraud, 0 = legitimate)

Obtaining the Data

1. Download the dataset from Kaggle
2. Place the creditcard.csv file in a data/ directory in your project folder

Installation & Setup

1. Clone the repository:

git clone https://github.com/yourusername/credit-card-fraud-dashboard.git

cd credit-card-fraud-dashboard

2. Install dependencies:

pip install streamlit pandas numpy plotly networkx matplotlib scikit-learn imbalanced-learn

3. Run the dashboard:

streamlit run app.py

Dashboard Features

• Client Selection: Choose between example fraud/legitimate clients or random clients
• Network Visualization: See Network Graph
• Detailed Transaction View: Explore all transactions with fraud highlighting
• Fraud Analytics: View fraud-specific metrics and timelines (View Dashboard Demo)