Source code for BEng Final year project: Data Analytics for Anomaly Detection in sensitive data access. Minh Nghia Nguyen (40143951) - Queen's University Belfast The machine learning model is defined in ocsvm.py. For more information please check https://github.com/minh-nghia/AE-1SVM Coordinates of countries and zones are retrieved using Geopy in convert_coor.py. The results are saved in zone_convert.json and country_convert.json (manually at the moment) JSON database is extracted into Numpy array data.npy in extract_database.py. Replace logs.json and alerts.json with proper file names. The dimensions of the extracted Numpy array are: 0. Principle index 1. Operation index 2. Resource index 3. Request IP index 4-6. Request country coordinates 7-9. Resource zone coordinates 10-13. Resource zone one-hot encoding of sub-zone (a, b, c, d, .etc) 14. Daytime cosine. 15. Daytime sine. 16. Weektime cosine. 17. Weektime sine. 18. 1 if allowed log else -1. tuning.py runs the model over a range of values for nu and gamma, with 5-fold cross validation saving.py is the main training code. UI.py implements the graphical interface using tkInter.