Multi-source Machine Learning for Cyber Security
Vision
We leverage model pruning and knowledge distillation for resource-efficient and privacy-preserving federated learning. Past research includes multi-source anomaly detection for advanced-persistent threat and insider threat detection.
Funded Projects
- Deep Resilience for Multifaceted Federated Learning in Internet-of-Everything – Commonwealth Cyber Initiative (CCI) 2021
- Content Analytics for Smart Search – Xerox Workplace Innovation Research 2018
- Advanced Persistent Threat Detection – Defense Advanced Research Projects Agency (DARPA) 2015
Publications
- Resource-Efficient Federated Learning for Heterogenous and Resource-Constrained Environments – Preprint 2023
- Computer-implemented system and method for detecting anomalies using sample-based rule identification – Patent Granted 2018
- Detecting anomalies in work practice data by combining multiple domains of information – Patent Granted 2016
- System and method for modeling behavior change and consistency to detect malicious insiders – Patent Granted 2015
- Method and apparatus for combining multi-dimensional fraud measurements for anomaly detection – Patent Granted 2014
- Multi-source fusion for anomaly detection: using across-domain and across-time peer-group consistency checks – Journal of Wireless Mobile Networks, Ubiquitous Computing, and Dependable Applications 2014
- Multi-domain information fusion for insider threat detection – IEEE Security and Privacy 2013
- Fraud Detection for Healthcare – Knowledge Discovery and Data Mining (KDD) Data Mining for Healthcare 2013
- Provenance Segmentation – USENIX Workshop on the Theory and Practice of Provenance (TaPP) 2016
- Diagnosing Advanced Persistent Threats: A Position Paper – Workshop on Principles of Diagnosis 2016
- Secure Two-Party Feature Selection – Preprint 2019