This project focuses on developing a Credit Card Fraud Detection system using machine learning techniques. The dataset used contains information about credit card transactions made by European cardholders in September 2013. The goal is to build models, specifically Decision Tree and Support Vector Machine (SVM), to predict whether a transaction is fraudulent or legitimate.
- Python
- Jupyter Notebook
- Scikit-Learn
- Snap ML
- Pandas
- NumPy
- Matplotlib
- Seaborn
- Time
- Hinge Loss Metric
- Implemented a Decision Tree model using Scikit-Learn.
- Developed a Support Vector Machine (SVM) model using Scikit-Learn.
- Utilized Snap ML for high-performance machine learning models.
- Data Preprocessing: Scaled and normalized features, performed data inflation.
- Model Building: Implemented Decision Tree and SVM models.
- Evaluation: Assessed model performance using ROC-AUC score and hinge loss metric.
- Trained models that demonstrate high accuracy in detecting fraudulent credit card transactions.
- Achieved consistency in model performance across different frameworks (Scikit-Learn and Snap ML).
- Showcase proficiency in machine learning model development.
- Provide a reference for implementing fraud detection systems.
- Demonstrate skills in preprocessing, model building, and evaluation.
- Highlight the ability to work with both Scikit-Learn and Snap ML libraries.
Feel free to fork, contribute, or use this project as a reference for your own fraud detection endeavors. Your feedback and contributions are highly appreciated!
