A Streamlit-based web app that detects potential credit card fraud using machine learning techniques on transaction datasets. It helps identify anomalous transactions based on behavioral patterns.
- Predicts whether a transaction is fraudulent or not.
- Trained on imbalanced dataset using machine learning (e.g., Logistic Regression, Random Forest, etc.).
- Visualizes correlations, distributions, and outliers.
- Real-time input interface for predictions.
- Dataset: Credit Card Fraud Detection Dataset - Kaggle
- Features are PCA-transformed due to confidentiality.
- Contains
284,807transactions, out of which492are frauds.
- Clone the repository
git clone https://github.com/yourusername/creditcard-fraud-detection.git
cd creditcard-fraud-detection