Menu
Anomaly Behavior Discovery
Advanced Machine Learning and Intelligent Decision Support Systems for User Anomaly Behavior Discovery in Large Transaction System
Our projects aim at assisting system administrators to enhance the security policies of their systems. In such environments, the authenticated and authorized users (trusted users) in one system can access to the resources of other systems and cause serious damage. The security administrators cannot effectively monitor and control such situations due to huge volumes of dynamically changing user behaviors and intentions.
Project Areas
We focus on various project types:
- Feature engineering to extract salient aspects of transactions that contribute to the system security provisioning
- User-behavior pattern language and constraint pattern matching to identify suspected behaviors
- User behavior simulation through generating large synthetic transaction datasets
These projects utilize artificial intelligence techniques (data mining and machine learning) through a high-performance computing planform including Apache Spark for high-speed in-memory computing and Nvidia DGX machine learning and GPU technologies.
Publications
UDATE PROJECT TITLE
B. Eapen, N. Archer, K. Sartipi | February 2021 | Journal of JMIR Dermatol 2020
UDATE PROJECT TITLE
B. Eapen, K. Sartipi, N. Archer | May 2020 | ArXiv.org
Research Team

Dr. Norm Archer
PROFESSOR EMERITUS
Information systems
DEGROOTE SCHOOL OF BUSINESS
MCMASTER UNIVERSITY
Other projects
Clinical Decision System
Medical Informatics
A multi-disciplinary research initiative on Information and System Intelligence research.