Cross-disciplinary research led as Assistant Professor at SUNY Polytechnic Institute (2018–2021) in collaboration with the SUNY Poly Computer Science group Research Group X∞ (Dr. Ali Tekeoglu, Dr. Sam Sengupta, and colleagues). Applied linear-system theory (Hankel-matrix methods, subspace identification, KL-divergence) to build anomaly-detection algorithms for cyber-physical systems, ICS/SCADA networks, and IoT devices. Code repositories for several of the papers below are hosted at the X∞ group page.

Anomaly detection — theory

• Hankel-based unsupervised anomaly detection — subspace-identification approach to time-series anomaly scoring (IEEE American Control Conference, 2020)
• Symmetric Kullback–Leibler divergence of softmaxed distributions for anomaly scores (IEEE Conf. on Communications and Network Security, 2019)
• Unsupervised time-series anomaly detection in ICS/SCADA networks (IEEE ISNCC 2021)
• Hybrid intrusion detection with machine learning in cloud computing environments (IEEE SERA 2019)

Applied fault detection

• Edge-based explainable fault detection for photovoltaic panels — on-edge inference for solar-panel fault diagnosis (Renewable Energy, 2022 — 62 citations; arXiv preprint)
• Proof-of-concept DoS attack against Bluetooth IoT devices — vulnerability analysis and detection (IEEE PerCom Workshop, 2020)

More on the group and paper code: Research Group X∞.