Research Program
My research develops trustworthy intelligent systems that continue to operate under uncertainty, distributed control, sensitive data, and adversarial pressure. Three connected programs carry this work from cybersecurity foundations to privacy-aware learning and biomedical applications.
1. Zero Trust and intelligent cyber defense
I study behavioral trust, context-aware anomaly detection, and adaptive defenses for systems whose topology, prior knowledge, and threat conditions cannot be assumed complete. This work includes Tenko, a distributed and lightweight intrusion-detection framework that converts packet-level anomaly evidence into persistent node-level trust scores, and extends toward agentic Zero Trust and deployable edge defense.
2. Privacy-preserving and trustworthy AI
I develop and evaluate methods for collaborative learning when sensitive information cannot be treated as freely available. This program spans federated and incremental learning, privacy attacks and defenses, federated unlearning, and robustness under changing data and threat conditions.
3. AI for biomedical and genomic research
I apply AI and machine learning to biomedical and genomic questions where data sensitivity, population differences, and cross-disciplinary interpretation shape the research problem. This program includes NIH AIM-AHEAD-funded work on prostate-cancer genomics and separate biomedical AI collaborations in privacy- and robustness-aware learning.
Across all three programs, the common thread is trustworthy intelligent systems operating with incomplete, distributed, sensitive, and adversarial information.
For the complete and current publication record, visit https://dblp.org/pid/05/5654-1