Research

My research interests are in statistical machine learning, deep learning theory, high-dimensional statistics, privacy-preserving data analysis, optimization, and large-scale multiple testing.

Papers

• Neurashed: A Phenomenological Model for Imitating Deep Learning Training
  Weijie Su. 2021.

• You Are the Best Reviewer of Your Own Papers: An Owner-Assisted Scoring Mechanism
  Weijie Su. Neural Information Processing Systems (NeurIPS), 2021. [supplementary][poster]

• Imitating Deep Learning Dynamics via Locally Elastic Stochastic Differential Equations
  Jiayao Zhang, Hua Wang, and Weijie Su. Neural Information Processing Systems (NeurIPS), 2021.

• An Unconstrained Layer-Peeled Perspective on Neural Collapse
  Wenlong Ji, Yiping Lu, Yiliang Zhang, Zhun Deng, and Weijie Su. 2021.

• Characterizing the SLOPE Trade-off: A Variational Perspective and the Donoho-Tanner Limit
  Zhiqi Bu, Jason Klusowski, Cynthia Rush, and Weijie Su. 2021.

• Weighted Training for Cross-Task Learning
  Shuxiao Chen, Koby Crammer, Hangfeng He, Dan Roth, and Weijie Su. 2021.

• A Central Limit Theorem for Differentially Private Query Answering
  Jinshuo Dong, Weijie Su, and Linjun Zhang. Neural Information Processing Systems (NeurIPS), 2021.

• A Theorem of the Alternative for Personalized Federated Learning
  Shuxiao Chen, Qinqing Zheng, Qi Long, and Weijie Su. 2021.

• Rejoinder: Gaussian Differential Privacy
  Jinshuo Dong, Aaron Roth, and Weijie Su. Journal of the Royal Statistical Society: Series B (Statistical Methodology), to appear, 2021.

• Exploring Deep Neural Networks via Layer-Peeled Model: Minority Collapse in Imbalanced Training
  Cong Fang, Hangfeng He, Qi Long, and Weijie Su. Proceedings of the National Academy of Sciences (direct submission), 118(43), 2021.

• Oneshot Differentially Private Top- Selection
  Gang Qiao, Weijie Su, and Li Zhang. International Conference on Machine Learning (ICML), 139, 8672–8681, 2021.

• Toward Better Generalization Bounds with Locally Elastic Stability
  Zhun Deng, Hangfeng He, and Weijie Su. International Conference on Machine Learning (ICML), 139, 2590–2600, 2021.

• Federated -Differential Privacy
  Qinqing Zheng, Shuxiao Chen, Qi Long, Weijie Su. International Conference on Artificial Intelligence and Statistics (AISTATS), 130, 2251–2259, 2021.

• Analysis of Information Transfer from Heterogeneous Sources via Precise High-dimensional Asymptotics
  Fan Yang, Hongyang Zhang, Sen Wu, Weijie Su, and Christopher Ré. 2020.

• Label-Aware Neural Tangent Kernel: Toward Better Generalization and Local Elasticity
  Shuxiao Chen, Hangfeng He, and Weijie Su. Neural Information Processing Systems (NeurIPS), 33, 15847–15858, 2020.

• Benign Overfitting and Noisy Features
  Zhu Li, Weijie Su, and Dino Sejdinovic. 2020.

• A Power Analysis for Knockoffs with the Lasso Coefficient-Difference Statistic
  Asaf Weinstein, Weijie Su, Malgorzata Bogdan, Rina Barber, and Emmanuel Candès. 2020.

• The Complete Lasso Tradeoff Diagram
  Hua Wang, Yachong Yang, Zhiqi Bu, and Weijie Su. Neural Information Processing Systems (NeurIPS), 33, 20051–20060, 2020.

• The Price of Competition: Effect Size Heterogeneity Matters in High Dimensions
  Hua Wang, Yachong Yang, and Weijie Su. 2020.

• On Learning Rates and Schrödinger Operators
  Bin Shi, Weijie Su, and Michael Jordan. 2020.

• Sharp Composition Bounds for Gaussian Differential Privacy via Edgeworth Expansion
  Qinqing Zheng, Jinshuo Dong, Qi Long, and Weijie Su. International Conference on Machine Learning (ICML), 119, 11420–11435, 2020.

• Towards Understanding the Dynamics of the First-Order Adversaries
  Zhun Deng, Hangfeng He, Jiaoyang Huang, and Weijie Su. International Conference on Machine Learning (ICML), 119, 2484–2493, 2020.

• Deep Learning with Gaussian Differential Privacy
  Zhiqi Bu, Jinshuo Dong, Qi Long, and Weijie Su. Harvard Data Science Review, 2020(23), 2020. [code]

• Robust Learning Rate Selection for Stochastic Optimization via Splitting Diagnostic
  Matteo Sordello, Hangfeng He, and Weijie Su. 2019.

• The Local Elasticity of Neural Networks
  Hangfeng He and Weijie Su. International Conference on Learning Representations (ICLR), 2020.

• Algorithmic Analysis and Statistical Estimation of SLOPE via Approximate Message Passing
  Zhiqi Bu, Jason Klusowski, Cynthia Rush, and Weijie Su. IEEE Transactions on Information Theory, 67(1), 506–537, 2020.

• Gaussian Differential Privacy
  Jinshuo Dong, Aaron Roth, and Weijie Su. Journal of the Royal Statistical Society: Series B (Statistical Methodology) (with Discussion), to appear, 2021.

• Acceleration via Symplectic Discretization of High-Resolution Differential Equations
  Bin Shi, Simon Du, Weijie Su, and Michael Jordan. Neural Information Processing Systems (NeurIPS), 32, 5744–5752, 2019.

• The FDR-Linking Theorem
  Weijie Su. 2018.

• Understanding the Acceleration Phenomenon via High-Resolution Differential Equations
  Bin Shi, Simon Du, Michael Jordan, and Weijie Su. Mathematical Programming, to appear, 2021.

• Differentially Private False Discovery Rate Control
  Cynthia Dwork, Weijie Su, and Li Zhang. Journal of Privacy and Confidentiality, 11(2), 2021.

• Robust Inference Under Heteroskedasticity via the Hadamard Estimator
  Edgar Dobriban and Weijie Su. 2018.

• HiGrad: Uncertainty Quantification for Online Learning and Stochastic Approximation
  Weijie Su and Yuancheng Zhu. 2018. [webpage][code]

• Detecting Multiple Replicating Signals using Adaptive Filtering Procedures
  Jingshu Wang, Lin Gui, Weijie Su, Chiara Sabatti, and Art Owen. The Annals of Statistics, to appear, 2021.

• Assumption Lean Regression
  Richard Berk, Andreas Buja, Lawrence Brown, Edward George, Arun Kuchibhotla, Weijie Su, and Linda Zhao. The American Statistician, 75(1), 76–84, 2021.

• Multiple Testing When Many p-Values Are Uniformly Conservative, with Application to Testing Qualitative Interaction in Educational Interventions
  Qingyuan Zhao, Dylan Small, and Weijie Su. Journal of the American Statistical Association, 114(527), 1291–1304, 2019.

• Statistical Inference for the Population Landscape via Moment Adjusted Stochastic Gradients
  Tengyuan Liang and Weijie Su. Journal of the Royal Statistical Society: Series B (Methodological), 81(2), 431–456, 2019.

• Group SLOPE – Adaptive Selection of Groups of Predictors
  Damian Brzyski, Alexej Gossmann, Weijie Su, and Małgorzata Bogdan. Journal of the American Statistical Association, 114(525), 419–433, 2019. [code]

• When Is the First Spurious Variable Selected by Sequential Regression Procedures?
  Weijie Su. Biometrika, 105 (3), 517–527, 2018.

• False Discoveries Occur Early on the Lasso Path
  Weijie Su, Małgorzata Bogdan, and Emmanuel Candès. The Annals of Statistics, 45(5), 2133–2150, 2017. [supplementary][code]

• A Differential Equation for Modeling Nesterov's Accelerated Gradient Method: Theory and Insights
  Weijie Su, Stephen Boyd, and Emmanuel Candès. Journal of Machine Learning Research, 17(153), 1–43, 2016. [arXiv][conference version][BibTex]

• SLOPE Is Adaptive to Unknown Sparsity and Asymptotically Minimax
  Weijie Su and Emmanuel Candès. The Annals of Statistics, 44(3), 1038–1068, 2016. [supplementary]

• Familywise Error Rate Control via Knockoffs
  Lucas Janson and Weijie Su. Electronic Journal of Statistics, 10(1), 960–975, 2016.

• SLOPE – Adaptive Variable Selection via Convex Optimization
  Małgorzata Bogdan, Ewout van den Berg, Chiara Sabatti, Weijie Su, and Emmanuel Candès. The Annals of Applied Statistics, 9(3), 1103–1140, 2015. [code]