Publications and Preprints

Deep Learning Theory

  • Bridging Unsupervised and Semi-Supervised Anomaly Detection: A Theoretically-Grounded and Practical Framework with Synthetic Anomalies (2025+).
    Tian-Yi Zhou*, Matthew Lau*, Xiangchi Yuan, Jizhou Chen, Wenke Lee, and Xiaoming Huo.
    Submitted.

  • Optimal Classification-based Anomaly Detection with Neural Networks: Theory and Practice in Cybersecurity (2024+).
    Tian-Yi Zhou*, Matthew Lau*, Jizhou Chen, Wenke Lee, and Xiaoming Huo.
    Under Revision at Annals of Applied Statistics (AOAS). ArXiv

  • Approximation of RKHS Functionals by Neural Networks (2024+).
    Tian-Yi Zhou, Namjoon Suh, Guang Cheng, and Xiaoming Huo.
    Under Revision at Journal of Machine Learning Research (JMLR). ArXiv

  • Classification of Unbounded Data generated by Gaussian Mixture Models via deep ReLU Networks (2024).
    Tian-Yi Zhou and Xiaoming Huo.
    Journal of Machine Learning Research (JMLR). Paper Link

  • Learning Ability of Interpolating Deep Convolutional Neural Networks (2023).
    Tian-Yi Zhou and Xiaoming Huo.
    Applied and Computational Harmonic Analysis (ACHA). Paper Link

  • Approximation and non-parametric estimation of functions over high-dimensional spheres via deep ReLU networks (2023).
    Namjoon Suh, Tian-Yi Zhou, and Xiaoming Huo.
    ICLR 2023. Paper Link

Computational Sociology

  • Drink like a Man? Modified Poisson Analysis of Adolescent Binge Drinking in the US, 1976–2022 (2024).
    Jiaxin Gu, Minheng Chen, Yue Yuan, Xin Guo, Tian-Yi Zhou, and Qiang Fu.
    Social Science & Medicine.Paper Link

  • Gone with the Weed: Incidents of Adolescent Marijuana Use in the United States, 1976-2021 (2023).
    Jiaxin Gu, Xin Guo, Xiaoxi Liu, Yue Yuan, Yushu Zhu, Minheng Chen, Tian-Yi Zhou, and Qiang Fu.
    Annals of Epidemiology. Paper Link

  • Modified Poisson regression analysis of grouped and right‐censored counts (2021).
    Qiang Fu, Tian-Yi Zhou, and Xin Guo.
    Journal of the Royal Statistical Society: Series A (Statistics in Society). Paper Link