How haroon423/fruad-detection-machine-learning works
File structure
Top‑level artifacts:
• .github/workflows/main.yml – CI/CD pipeline (runs tests)
• requirements.txt – Python dependencies
• test_fraud_detection.py – unit/integration tests for core functionality
• README.md – project overview
• Fruad-Detection-Using-Machine_learning/ – core notebooks, model outputs, dissertation
Inside “Fruad-Detection-Using-Machine_learning”:
• Algorithms.ipynb – model‑training code for traditional ML
• Data_Visualisation.ipynb – visualization code for performance and features
• Dissertation.pdf – thesis write‑up (contextual document)
• Models/Scaled/… – stored .h5 model files and loss logs for RNN and LSTM under “More Epochs” and “Suggestion” experiments
README for additional insights
• Sections on data preprocessing, feature engineering, model training, and evaluation
• Lists ML algorithms and performance metrics (precision, recall, F1, ROC‑AUC)
• Instructions for setup and running the pipeline locally
Detailed capture of architectural elements
• Under Training, sub‑blocks for “Random Forest”, “Gradient Boosting”, “SVM” and “RNN/LSTM”
• Under Models, illustrate directory structure:
Models → Scaled → More Epochs / Suggestion → {Model 1, Model 2}
• How test_fraud_detection.py fits into CI/CD before merging