Biography

I am a tenure-track faculty at CISPA Helmholtz Center for Information Security. I obtained my Ph.D degree in computer science at University of Virginia, advised by Prof. David Evans. I received my M.S. degree from Department of Statistics at University of Virginia and my B.S. degree in Mathematics at Tsinghua University. My research interests cover a variety of topics in machine learning and security, including trustworthy machine learning, adversarial machine learning, statistical machine learning, convex/non-convex optimization and deep learning. Here is my latest CV.

I am looking for strong and motivated students who are interested in trustworthy machine learing. Send me an email if you are interested.
Interests
  • Trustworthy Machine Learning
  • Adversarial Machine Learning
  • Statistical Machine Learning
  • Optimization
Education

  • PhD in Computer Science, Aug. 2017 - Aug. 2022

    University of Virginia

  • MS in Statistics, Aug. 2015 - May 2017

    University of Virginia

  • BS in Mathematics, Sep. 2011 - Jul. 2015

    Tsinghua University

Recent Publications

Quickly discover relevant content by filtering publications.
Xiao Zhang, David Evans (2021). Incorporating Label Uncertainty in Intrinsic Robustness Measures. ICLR 2021 aisecure workshop.

PDF Poster Link

Jack Prescott, Xiao Zhang, David Evans (2021). Improved Estimation of Concentration Under Lp-Norm Distance Metric Using Half Spaces. ICLR 2021.

PDF Cite Code ArXiv OpenReview Post

Xiao Zhang, Jinghui Chen, Quanquan Gu, David Evans (2020). Understanding the intrinsic robustness of image distributions using conditional generative models. AISTATS 2020.

PDF Cite Code Slides ArXiv Link

Sicheng Zhu, Xiao Zhang, David Evans (2020). Learning Adversarially Robust Representations via Worst-Case Mutual Information Maximization. ICML 2020.

PDF Cite Code ArXiv Link Post

Xiao Zhang, Yaodong Yu, Lingxiao Wang, Quanquan Gu (2019). Learning One-hidden-layer ReLU Networks via Gradient Descent. AISTATS 2019.

PDF Cite Poster ArXiv Link

See all publications

Recent Talks

Empirically Measuring Concentration: Fundamental Limits to Intrinsic Robustness
We propose a novel method to measure the concentration of any distributions using empirical samples and show that concentration of measure does not prohibit the existence of adversarially robust classifiers for several image benchmarks.
Dec 10, 2019 10:35 AM — 10:40 AM Vancouver Convention Center
Saeed Mahloujifar, Xiao Zhang, Mohammad Mahmoody, David Evans
PDF Code Slides Video
Empirically Measuring Concentration: Fundamental Limits to Intrinsic Robustness
Fast and Sample Efficient Inductive Matrix Completion via Multi-Phase Procrustes Flow
We present a new gradient-based optimization algorithm for inductive matrix completion, which achieves both linear rate of convengence and sample complexities linearly depending on the feature dimension.
Jul 13, 2018 8:20 AM — 8:30 AM Stockholmsmässan
Xiao Zhang, Simon Du, Quanquan Gu
PDF Code Slides Video
A Unified Computational and Statistical Framework for Nonconvex Low-Rank Matrix Estimation
A generic framework based on a new stochastic variance-reduced gradient descent algorithm for accelerating nonconvex low-rank matrix recovery
Aug 7, 2017 12:51 PM — 1:09 AM International Convention Centre
Lingxiao Wang, Xiao Zhang, Quanquan Gu
PDF Code Slides Video

Internship Experience

 
 
 
 
 
Machine Learning Research Intern
Robert Bosch LLC
Jun 2020 – Oct 2020 Pittsburgh
Building efficient adversarially robust classification models using meta learning techniques (mentor: Anit Kumar Sahu)

Teaching Experience

Teaching Assistent - University of Virginia

Department of Computer Science

  • CS3102: Theory of Computation (Fall ‘19)
  • CS6501: Optimization for Machine Learning – Teaching Assistant (Fall ‘17)
  • CS2102: Discrete Math – Teaching Assistant (Fall ‘17)

Teaching Assistent - University of Virginia

Department of Statistics

  • STAT2120: Introduction to Statistical Science (Fall ‘16, Spring ‘17)

Professional Services

Program Committee member

  • IEEE Conference on Secure and Trustworthy Machine Learning (SaTML)
  • IEEE European Symposium on Security and Privacy (Euro S&P)

Journal Reviewer

  • Machine Learning (MLJ)
  • Advances in Computational Mathematics (ACOM)
  • Journal of Intelligent Information Systems (JIIS)
  • Transactions on Machine Learning Research (TMLR)

Conference Reviewer

  • Neural Information Processing Systems (NeurIPS)
  • International Conference on Machine Learning (ICML)
  • International Conference on Learning Representations (ICLR)
  • International Conference on Artificial Intelligence and Statistics (AISTATS)

Popular Topics

Adversarial Examples Algorithm Certified Robustness Concentration of Measure Cost-Sensitive Learning Deep Learning Generative Models Inductive Matrix Completion Intrinsic Robustness Label Uncertainty Low-Rank Matrix Recovery Mutal Information Nonconvex Optimization Phase Retrieval Primal-Dual Analysis ReLU Activation Representation Learning Variance Reduction

Contact