Overview
My research in machine learning and AI focuses on developing scalable and provable algorithms, optimization methods and theory, reinforcement learning techniques, and foundation models with applications in health, science, and engineering. These contributions have appeared in leading conferences such as ICML, ICLR, and NeurIPS, as well as flagship IEEE and INFORMS journals.

Selected Publications in Machine Learning and AI
(group members are underlined and corresponding authors are denoted by *)
  • Yu, X., Wang, Y., Chen, J., and Xue, L.* (2025)
    AltLoRA: Towards Better Gradient Approximation in Low-Rank Adaptation with Alternating Projections (arXiv)
    The Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS).
  • Zhang, H., Zheng, Z. (co-first author), and Xue, L.* (2025)
    Regret-Optimal Q-Learning with Low Cost for Single-Agent and Federated Reinforcement Learning. (arXiv)
    The Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS).
  • Li, S., Ding, Y., Xue, L.*, and Li, R. (2025)
    Stability and Oracle Inequalities for Optimal Transport Maps between General Distributions
    The Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS).
  • Yang, H., Zhang, T. (co-first author), and Xue, L.* (2025)
    Theoretical Guarantees for Sparse Principal Component Analysis based on the Elastic Net. (link, arXiv)
    IEEE Transactions on Information Theory, in press.
  • Zhang, H., Zheng, Z. (co-first author), and Xue, L.* (2025)
    Gap-Dependent Bounds for Federated Q-Learning. (openreview, arXiv)
    The Forty-Second International Conference on Machine Learning (ICML).
  • Yu, X., He, Z. (co-first author), Sun, Y.*, Xue, L.*, and Li, R. (2025)
    Understanding the Statistical Accuracy-Communication Trade-off in Personalized Federated Learning with Minimax Guarantees. (openreview, arXiv)
    The Forty-Second International Conference on Machine Learning (ICML).
  • Zheng, Z., Zhang, H. (co-first author) and Xue, L.* (2025)
    Gap-Dependent Bounds for Q-Learning using Reference-Advantage Decomposition. (openreview, arXiv)
    The Thirteenth International Conference on Learning Representations (ICLR). (Spotlight)
  • Zheng, Z., Zhang, H. (co-first author) and Xue, L.* (2025)
    Federated Q-Learning with Reference-Advantage Decomposition: Almost Optimal Regret and Logarithmic Communication Cost. (openreview, arXiv)
    The Thirteenth International Conference on Learning Representations (ICLR).
  • Zheng, Z., Gao, F., Xue, L.* and Yang, J.* (2024)
    Federated Q-Learning: Linear Regret Speedup with Low Communication Cost. (openreview, arXiv)
    The Twelfth International Conference on Learning Representations (ICLR).
  • Zheng, Z., Ma, S.* and Xue, L.* (2024)
    A New Inexact Proximal Linear Algorithm with Adaptive Stopping Criteria for Robust Phase Retrieval. (link, arXiv)
    IEEE Transactions on Signal Processing, 72: 1081-1093.
  • Wang, Z., Liu, B. (co-first author), Chen, S., Ma, S., Xue, L.* and Zhao, H. (2022)
    A Manifold Proximal Linear Method for Sparse Spectral Clustering with Application to Single-Cell RNA Sequencing Data Analysis. (link, arXiv)
    INFORMS Journal on Optimization, 4: 200-214.
  • Wang, B., Ma, S. and Xue, L.. (2022)
    Riemannian Stochastic Proximal Gradient Methods for Nonsmooth Optimization over Stiefel Manifold. (link, arXiv)
    Journal of Machine Learning Research, 23: 1−33.
  • Chen, S., Ma, S., Xue, L. and Zou, H. (2020)
    An Alternating Manifold Proximal Gradient Method for Sparse PCA and Sparse CCA. (link, arXiv)
    INFORMS Journal on Optimization, 2: 192-208.
  • Zou, H. and Xue, L. (2018)
    A Selective Overview of Sparse Principal Component Analysis. (link)
    Proceedings of the IEEE, 106: 1311-1320.