- Modeled past credit card transactions with verified fraud transactions and identified whether the new transactions are fraud.
- Utilized Python and Jupyter notebook for visualization and analysis.
- Utilized Isolation Forest Algorithm and Local Outlier Factor (LOF) Algorithm to achieve 99.4% accuracy using Isolation Forest while achieving 99.65% accuracy with LOF.
Data Source: https://www.kaggle.com/mlg-ulb/creditcardfraud