About Ian

I am a tenure-track assistant professor at CS@UMass Lowell. I work on machine learning and IoT. Before joining UML, I was a postdoc at CNSR@VT and obtained my Ph.D. at CNSG@ASU. I received my B.S. from EECS@Peking University and M.Phil from EE@The Chinese University of Hong Kong.

  • We have ongoing Colloquium at CS@UMass Lowell. Check out the exciting talks here!


  • Our paper Squeezing More Utility via Adaptive Clipping on Deferentially Private Gradients in Federated Meta-Learning is accepted by ACSAC 2022 (Acceptance ratio: 73/303=24.0%). The idea is to adapt differential privacy to address data privacy challenges in Federated Meta-Learning. Congrutulation to Ning and other authors. (Sep. 2022)
  • Our paper Clang __usercall: Towards Native Support for User Defined Calling Conventions is accepted by ACM ESEC/FSE 2022 Demo track. The idea is to mimic popular syntax and adapting Clang for interfacing purpose of C/C++ code. Congrutulation to Jared and other authors. (Aug. 2022)
  • Our paper Transferability of Adversarial Examples in Machine Learning-based Malware Detection is accepted by IEEE CNS 2022. The idea is to spread out adversarial perturbations in order to improve transferability of adversary example attacks. Congrutulation to Yang and other authors. (Aug. 2022)
  • Ian is to serve as a TPC member for 2022 IEEE INFOCOM, CNS, and MASS. (Apr. 2022)
  • Our paper FLARE: Defending Federated Learning against Model Poisoning Attacks via Latent Space Representations is accepted by ACM ASIACCS 2022 (Acceptance ratio: 18.4%). The idea is to detect model poisoning attacks on federated learning by exploring the penultimate layer representations of neural network. Congrutulation to Ning and other authors. (Feb. 2022)
  • Our paper MANDA: On Adversarial Example Detection for Network Intrusion Detection System is accepted by IEEE TDSC. The idea is to detect adversarial example attacks by exploring the data space. Congrutulation to Ning and other authors. (Feb. 2022)
  • Our paper FeCo: Boosting Intrusion Detection Capability in IoT Networks via Contrastive Learning has been accepted by IEEE INFOCOM 2022 (Acceptance ratio: 225/1129=19.9%). The idea is to use contrastive learning to learn the representation of benign traffic. Congrutulation to Ning and other authors. (Dec. 2021)
  • Ian joined CS@UML as a TTAP. (Sep. 2021)