Portrait of Yue Wu

Yue Wu 吴悦

Postdoctoral Researcher

Department of Biostatistics, Epidemiology and Informatics
University of Pennsylvania

Google Scholar

I am a postdoctoral researcher at the University of Pennsylvania, working with Professor Yong Chen. I received my Ph.D. in Mathematics from the Hong Kong University of Science and Technology in 2023, advised by Professors Yang Xiang and Ke Wang.

My research develops theory and methods for statistical learning and scientific computing, with an emphasis on reliability and computational efficiency. I design and analyze neural function representations tailored to the mathematical structure and practical constraints of each problem. My recent work focuses on learning from distributed, heterogeneous, and imperfect data. This work spans federated inference, transfer learning, and AI methods for data analysis, with applications to biomedical research.

Selected Publications & Manuscripts

* Corresponding author. † Equal contribution.

Statistical methodology & applications

Federated Likelihood-based Inference via Sobolev Neural Approximation

Yue Wu, Huiyuan Wang, Runze Li, Jianqing Fan, and Yong Chen

Manuscript, 2026. Under revision.

Neural representations of local likelihoods for one-shot federated estimation and inference without pooling individual-level data.

Manuscript available upon request

Low-Complexity Adjustment for Transfer Learning in Nonparametric Regression

Yue Wu, Jie Hu, Runze Li, and Yong Chen

Manuscript, 2026. Under review.

Leveraging low-complexity source-to-target adjustments to improve nonparametric regression across heterogeneous domains.

Manuscript available upon request

UNO: Neural Pruning Enhances AI Reliability of Observational Data-based Findings

Yue Wu†, Yiwen Lu†, Huiyuan Wang, Dazheng Zhang, Huilin Tang, Bingyu Zhang, Lu Li, Wenjie Hu, Jiajie Chen, Tingyin Wang, Han Wu, Ting Zhou, Yuqing Lei, Xinyao Jian, Tsai Hor Chan, Siqi Chen, Yong Lin, Demissie Alemayehu, James Zou, David Asch, and Yong Chen

Manuscript, 2026. Under revision.

Using negative controls to guide neural pruning and fine-tuning for observational data analysis.

Manuscript available upon request Code

TrialOptimizer: Multi-Objective Optimization of Clinical Trial Eligibility with Real-World Data

Yue Wu, Bingyu Zhang, Yiwen Lu, Jingyue Huang, Huiyuan Wang, Huilin Tang, and Yong Chen

Manuscript, 2026. Under revision.

A data-driven framework for evaluating trade-offs in clinical trial eligibility criteria using real-world data.

Manuscript available upon request Code

Scientific computing & neural network approximation

Score-based transport modeling for mean-field Fokker–Planck equations

Jianfeng Lu, Yue Wu*, and Yang Xiang

Journal of Computational Physics, 503:112859, 2024

Paper

Feature flow regularization: Improving structured sparsity in deep neural networks

Yue Wu, Yuan Lan, Luchan Zhang, and Yang Xiang

Neural Networks, 161:598–613, 2023

Paper

A Convergent ADMM Algorithm for Grain Boundary Energy Minimization

Yue Wu, Luchan Zhang, and Yang Xiang

Journal of Scientific Computing, 2026

Paper

Why does the two-timescale Q-learning converge to different mean field solutions? A unified convergence analysis

Jing An, Jianfeng Lu, Yue Wu, and Yang Xiang

Accepted for publication in Multiscale Modeling and Simulation, 2026

arXiv

Nearly optimal approximation rates for deep super ReLU networks on Sobolev spaces

Yahong Yang, Yue Wu, Haizhao Yang, and Yang Xiang

Preprint, 2023. Under revision.

arXiv

Selected Talks

Teaching

Teaching Assistant, Hong Kong University of Science and Technology

Teaching interests: Calculus, linear algebra, statistical learning, and machine learning.

Service

Journal reviewer: SIAM Journal on Numerical Analysis.
Session chair: First HKUST SIAM Student Chapter Annual Meeting, June 2023.