Pan Xu

Home | Publications


I am a Ph.D. student in the Department of Computer Science at the University of California, Los Angeles, supervised by Dr. Quanquan Gu. Before this, I was a Ph.D. student in Department of Computer Science at University of Virginia. I received my Bachelor's degree in Probability and Statistics from University of Science and Technology of China. My research interests include Machine Learning, Optimization and High Dimensional Statistics.

EDUCATION

  • Ph.D. in Computer Science (2018-now)
    Department of Computer Science
    University of California, Los Angeles

  • Ph.D. in Computer Science (2017-2018)
    Department of Computer Science
    University of Virginia

  • Ph.D. Candidate in Systems and Information Engineering (2015-2017)
    Department of Systems and Information Engineering
    University of Virginia

  • B.S. in Probability and Statistics (2011-2015)
    School of Mathematical Sciences
    University of Science and Technology of China



Email: panxu(at)cs(dot)ucla(dot)edu
Address: EVI 295, 404 Westwood Plaza, Los Angeles, CA 90095
  • Sample Efficient Stochastic Variance-Reduced Cubic Regularization Method [ArXiv]
    Dongruo Zhou, Pan Xu, Quanquan Gu, arXiv:1811.11989, 2018.

  • Finding Local Minima via Stochastic Nested Variance Reduction [ArXiv]
    Dongruo Zhou, Pan Xu, Quanquan Gu, arXiv:1806.08782, 2018.

  • Communication-efficient Distributed Estimation and Inference for Transelliptical Graphical Models [ArXiv]
    Pan Xu, Lu Tian, Quanquan Gu, arXiv:1612.09297, 2016.

Conference Proceedings

  • Global Convergence of Langevin Dynamics Based Algorithms for Nonconvex Optimization [Link] [ArXiv][Video]
    Pan Xu*, Jinghui Chen*, Difan Zou, Quanquan Gu (*: equal contribution)
    In Proc. of the 31st Conference on Advances in Neural Information Processing Systems (NeurIPS'2018), Montréal, Canada, 2018. [Spotlight, 3.5%]

  • Stochastic Nested Variance Reduction for Nonconvex Optimization [Link] [ArXiv][Video]
    Dongrou Zhou, Pan Xu, Quanquan Gu
    In Proc. of the 31st Conference on Advances in Neural Information Processing Systems (NeurIPS'2018), Montréal, Canada, 2018. [Spotlight, 3.5%]

  • Third-order Smoothness Helps: Even Faster Stochastic Optimization Algorithms for Finding Local Minima [Link] [ArXiv][Video]
    Yaodong Yu*, Pan Xu*, Quanquan Gu (*: equal contribution)
    In Proc. of the 31st Conference on Advances in Neural Information Processing Systems (NeurIPS'2018), Montréal, Canada, 2018.

  • Subsampled Stochastic Variance-Reduced Gradient Langevin Dynamics [Link]
    Difan Zou*, Pan Xu*, Quanquan Gu (*: equal contribution)
    In Proc. of the 34th International Conference on Uncertainty in Artificial Intelligence (UAI'2018), Monterey, California, 2018.

  • Continuous and Discrete-Time Accelerated Stochastic Mirror Descent for Strongly Convex Functions [Link]
    Pan Xu*, Tianhao Wang*, Quanquan Gu (*: equal contribution)
    In Proc. of the 35th International Conference on Machine Learning (ICML'2018), Stockholm, Sweden, 2018.

  • Stochastic Variance-Reduced Cubic Regularized Newton Method [Link][ArXiv]
    Dongruo Zhou, Pan Xu, Quanquan Gu
    In Proc. of the 35th International Conference on Machine Learning (ICML'2018), Stockholm, Sweden, 2018.

  • Stochastic Variance-Reduced Hamilton Monte Carlo Methods [Link][ArXiv]
    Difan Zou*, Pan Xu*, Quanquan Gu (*: equal contribution)
    In Proc. of the 35th International Conference on Machine Learning (ICML'2018), Stockholm, Sweden, 2018.

  • Covariate Adjusted Precision Matrix Estimation via Nonconvex Optimization [Link]
    Jinghui Chen, Pan Xu, Lingxiao Wang, Jian Ma, Quanquan Gu
    In Proc. of the 35th International Conference on Machine Learning (ICML'2018), Stockholm, Sweden, 2018.

  • Accelerated Stochastic Mirror Descent: From Continuous-time Dynamics to Discrete-time Algorithms [Link][PDF]
    Pan Xu*, Tianhao Wang*, Quanquan Gu (*: equal contribution)
    In Proc. of the 21st International Conference on Artificial Intelligence and Statistics (AISTATS'2018), Playa Blanca, Lanzarote, Canary Islands, 2018.

  • Speeding Up Latent Variable Gaussian Graphical Model Estimation via Nonconvex Optimization [Link][ArXiv][Poster][Code]
    Pan Xu, Jian Ma, Quanquan Gu
    In Proc. of the 30th Conference on Advances in Neural Information Processing Systems (NIPS'2017), Long Beach, CA, USA, 2017.

  • Uncertainty Assessment and False Discovery Rate Control in High-Dimensional Granger Causal Inference [Link]
    Aditya Chaudhry, Pan Xu, Quanquan Gu
    In Proc. of the 34th International Conference on Machine Learning (ICML'2017), Sydney, Australia, 2017.

  • Efficient Algorithm for Sparse Tensor-variate Gaussian Graphical Models via Gradient Descent [Link]
    Pan Xu, Tingting Zhang, Quanquan Gu
    In Proc. of the 20th International Conference on Artificial Intelligence and Statistics (AISTATS'2017), Fort Lauderdale, Florida, USA, 2017.

  • Semiparametric Differential Graph Models [Link][Video]
    Pan Xu, Quanquan Gu
    In Proc. of the 29th Conference on Advances in Neural Information Processing Systems (NIPS'2016), Barcelona, Spain, 2016.

  • Forward Backward Greedy Algorithms for Multi-Task Learning with Faster Rates [Link]
    Lu Tian, Pan Xu, Quanquan Gu
    In Proc. of the 32nd International Conference on Uncertainty in Artificial Intelligence (UAI'2016), New York / New Jersey, USA, 2016.