Yuxin Chen

I am an associate professor of Statistics and Data Science and of Electrical and Systems Engineering at University of Pennsylvania. Before moving to UPenn in 2022, I was an assistant professor of Electrical and Computer Engineering, and an associated faculty member of Computer Science and Applied and Computational Mathematics at Princeton University from 2017 to 2021. Prior to this, I was a postdoc in Statistics at Stanford University from 2015 to 2017, and obtained my Ph.D. in Electrical Engineering at Stanford University in Jan 2015. Research areas: mathematical data science, statistics, information theory, optimization, reinforcement learning, and their applications to medical imaging, power electronics, and computational biology. Contact: 313 Academic Research Building 265 South 37th Street, Philadelphia, PA 19104 Email: yuxinc at wharton dot upenn dot edu

Openings

I'm looking for highly motivated postdocs and Ph. D. students with strong mathematical background and interest in machine learning theory (particularly reinforcement learning), statistics, and optimization.

Recent news

Yuling Yan will join University of Wisconsin–Madison as an Assistant Professor of Statistics in 2024, after being a Norbert Wiener postdoc at MIT (advised by P. Rigollet and M. Wainwright). Congratulations, Yuling!! Yanxi has won the 2022 ICML outstanding paper award for his mixed LDS paper. Congratulations, Yanxi! Yuling has been awarded the prestigious Princeton Honorific Fellowship, 2022. Congratulations, Yuling! I received the 2022 Alfred P. Sloan Research Fellowship. Our new monograph "Spectral Methods for Data Science: A Statistical Perspective" is published in Foundations and Trends in Machine Learning. The updated version includes new sections on distributional theory and uncertainty quantification for spectral methods.

Topic courses I have developed

• Spring 2022: STAT991-302: Mathematics of High-Dimensional Data (UPenn)
  

• Fall 2020: ELE520: Mathematics of Data Science (Princeton)
  

• Fall 2019: ELE522: Large-Scale Optimization for Data Science (Princeton)
  

Selected recent papers

• Reinforcement learning

  • G. Li, Y. Chi, Y. Wei, Y. Chen, “Minimax-optimal multi-agent RL in Markov games with a generative model,” NeurIPS 2022 (selected as oral).

  • G. Li, L. Shi, Y. Chen, Y. Chi, Y. Wei, “Settling the sample complexity of model-based offline reinforcement learning,” 2022.

  • G. Li, Y. Wei, Y. Chi, Y. Chen, “Softmax policy gradient methods can take exponential time to converge,” accepted to Mathematical Programming, 2023+ (appeared in part to COLT 2021). [slides]

  • G. Li, C. Cai, Y. Chen, Y. Wei, Y. Chi, “Is Q-Learning minimax optimal? A tight sample complexity analysis,” accepted to Operations Research, 2023+ (appeared in part to ICML 2021).

  • S. Cen, C. Cheng, Y. Chen, Y. Wei, Y. Chi, “Fast global convergence of natural policy gradient methods with entropy regularization,” Operations Research, vol. 70, no. 4, pp. 2563–2578, 2022 (INFORMS George Nicholson award finalist, 2021). [ArXiv][OR version][slides]

  • G. Li, Y. Wei, Y. Chi, Y. Chen, “Breaking the sample size barrier in model-based reinforcement learning with a generative model,” accepted to Operations Research, 2023+ (appeared in part to NeurIPS 2020). [paper][slides]

• Nonconvex statistical learning and estimation

  • Y. Chen, J. Fan, C. Ma, Y. Yan, “Bridging convex and nonconvex optimization in robust PCA: Noise, outliers, and missing data,” Annals of Statistics, vol. 49, no. 5, pp. 2948-2971, Oct. 2021. [ArXiv][AoS version]

  • C. Cai, G. Li, H. V. Poor, Y. Chen, “Nonconvex low-rank tensor completion from noisy data,” Operations Research, vol. 70, no. 2, pp. 1219–1237, 2022 (appeared in part in NeurIPS 2019). [ArXiv][OR version][slides]

  • Y. Chen, J. Fan, C. Ma, Y. Yan, “Inference and uncertainty quantification for noisy matrix completion,” Proceedings of the National Academy of Sciences (PNAS), vol. 116, no. 46, pp. 22931–22937, Nov. 2019. [ArXiv][PNAS version][slides]

  • Y. Chen, Y. Chi, J. Fan, C. Ma, Y. Yan, “Noisy matrix completion: Understanding statistical guarantees for convex relaxation via nonconvex optimization,” SIAM Journal on Optimization, vol. 30, no. 4, pp. 3098–3121, October 2020. [ArXiv][slides]

  • Y. Chen, Y. Chi, J. Fan, C. Ma, “Gradient descent with random initialization: Fast global convergence for nonconvex phase retrieval,” Mathematical Programming, vol. 176, no. 1-2, pp. 5-37, July 2019. [ArXiv][slides]

  • C. Ma, K. Wang, Y. Chi, Y. Chen, “Implicit regularization in nonconvex statistical estimation: Gradient descent converges linearly for phase retrieval, matrix completion, and blind deconvolution,” Foundations of Computational Mathematics, vol. 20, no. 3, pp. 451-632, June 2020 (appeared in part in ICML 2018). [main text][supplement][full paper (Arxiv)][slides]

  • Y. Chen and E. J. Candes, “Solving random quadratic systems of equations is nearly as easy as solving linear systems,” Communications on Pure and Applied Mathematics, vol. 70, issue 5, pp. 822-883, May 2017. [slides][website]

• Spectral methods

  • Y. Zhou, Y. Chen, “Deflated HeteroPCA: Overcoming the curse of ill-conditioning in heteroskedastic PCA,” 2023. [ArXiv]

  • Y. Yan, Y. Chen, J. Fan, “Inference for heteroskedastic PCA with missing data,” 2021. [ArXiv]

  • C. Cai, G. Li, Y. Chi, H. V. Poor, Y. Chen, “Subspace estimation from unbalanced and incomplete data matrices, $\ell_{2,\infty}$ statistical guarantees ” Annals of Statistics, vol. 49, no. 2, pp. 944-967, 2021. [ArXiv][AoS version]

  • Y. Chen, C. Cheng, J. Fan, “Asymmetry helps: Eigenvalue and eigenvector analyses of asymmetrically perturbed low-rank matrices,” Annals of Statistics, vol. 49, no. 1, pp. 435-458, 2021. [ArXiv][AoS version][slides]

  • Y. Chen, J. Fan, C. Ma, K. Wang, “Spectral method and regularized MLE are both optimal for top-K ranking,” Annals of Statistics, vol. 47, no. 4, pp. 2204-2235, August 2019. [ArXiv][AoS version][slides]