Department of Computer Science

University of Illinois at Urbana-Champaign

4310 Siebel Center 201 N. Goodwin Ave.

Urbana, IL 61801, USA

Email:    lxbosky at gmail dot com

Or            lbo at illinois dot edu

​Office:    4310 Siebel

Assistant Professor

Research Interests

Recent News

  • We will organize a workshop "Security and Privacy of Machine Learning" in ICML 2019. Please submit your papers here and win the best paper award!
  • ​We will organize a workshop "Adversarial Machine Learning in Real-World Computer Vision Systems" in CVPR 2019. Please submit your papers here!
  • ​Our paper "Realistic Adversarial Examples in 3D Meshes" is accepted in CVPR 2019 as oral presentation! Congratulations to Chaowei and Dawei!
  • Our paper "Generating 3D Adversarial Point Clouds" is accepted in CVPR 2019! 
  • ​Our paper "How You Act Tells a Lot: Privacy-Leakage Attack on Deep Reinforcement Learning" got accepted in AAMAS 2019 as oral presentation!
  • ​Our paper "Towards Efficient Data Valuation Based on the Shapley Value" got accepted in AISTATS 2019! Check it out if you want to know which data contribute more to your model!
  • Our paper "Characterizing Audio Adversarial Examples Using Temporal Dependency" got accepted in ICLR 2019.

Bo Li

I am interested in machine learning, security, privacy, game theory, blockchain and related topics. I have designed several robust learning algorithms, a scalable framework for achieving robustness for a range of learning methods, and a privacy preserving data publishing system. I am currently working on anomaly detection systems against causative poisoning attacks and malware detection with real world collected big data. I'm also working on adversarial deep learning for training generative adversarial networks (GAN) and designing robust deep neural networks against adversarial attacks. Theoretically, I utilize game theoretic analyses to model the interactions between an intelligent adversary and a machine learner, allowing defender to design robust learning strategies that explicitly account for an adversary’s optimal response. Empirically, my current research aims to scalable robust algorithms that can process massive amounts of data available for Internet-scale problems regarding specific cloud computing infrastructure to achieve large-scale secure learning for big data.