Credit Card Fraud Detection

Datasets:

• card_transdata (kaggle)
• anonymous_data (kaggle)

Transformations:

• standardization
• balancing through oversampling libraries
• feature engineering

Models

• Logistic Regression
• Decision Tree
• XGBoost
• Random Forest Classifier
• Isolation Forest
• Support Vector Machine
• Muli-layer Perceptron Classifier
• KMeans clustering
• HAC

Setup

Install Project Dependencies

conda create --name fraud-detection python=3.10
conda activate fraud-detection
pip install -r requirements.txt

Report here

Metrics