Description
This Python-based capstone project is a complete machine learning system that detects fraudulent transactions using a synthetic dataset. It includes data cleaning, model training (Random Forest), and a basic GUI interface for prediction. CSV file parsing and preprocessing are handled internally, and model performance was evaluated with custom testing tools.
Project Files
Python Source Code
- main.py
- predict_fraud.py
- user_gui.py
- clean_csv.py
- random_forest.py
- testing.py
Dataset
- synthetic_fraud_dataset.csv
Click a file to view its contents here...
Capstone Report
This report documents the full development lifecycle of a fraud detection system created for a fictional company, Apex Banking. It covers problem identification, planning, data engineering, machine learning model development, GUI/CLI integration, and testing. The system uses Python and a Random Forest classifier to detect credit card fraud within a synthetic dataset.
- Designed for fictional company Apex Banking to detect credit card fraud
- Achieved 81.7% accuracy using a Random Forest classifier
- Built both CLI and GUI interfaces for end-user interaction
- Processed and balanced transaction data from a synthetic CSV dataset
You can view the full report here (PDF, 18 pages).
What I Learned
This project reinforced my skills in machine learning workflows, especially data preprocessing, feature engineering, and model evaluation. I also learned how to integrate a GUI interface with backend predictions, and how to manage real-world file input using Python.