​​

Statistical Machine Learning

[ODE/PDE Models, Multimodal Learning, Kernel-Based Learning]

(Note:     students under my supervision;  † equal contribution;  * corresponding author)​​​​

​

ODE/PDE Models

​

• On large batch training and sharp minima: A Fokker-Planck perspective. [journal][reprint][preprint]​[older version]

Xiaowu Dai and Yuhua Zhu. 

Journal of Statistical Theory and Practice (JSTP), special issue on "Advances in Deep Learning", 2020.

​

• Kernel ordinary differential equations. [journal][reprint][preprint]

Xiaowu Dai and Lexin Li.​ 

Journal of the American Statistical Association (JASA), 2022.​

​

• Post-regularization confidence bands for ordinary differential equations. [journal][preprint]

Xiaowu Dai​* and Lexin Li. 

Journal of Machine Learning Research (JMLR), 2024.​

​

• Nonparametric estimation via partial derivatives. [journal​][reprint​][preprint]

Xiaowu Dai*. 

Journal of the Royal Statistical Society Series B: Statistical Methodology (JRSSB), 2025. 

​

• Quantifying microbial interactions based on compositional data using an iterative approach for solving generalized Lotka-Volterra equations. [journal][reprint​][code​]

Yue Huang, Tianqi Tang, Xiaowu Dai, and Fengzhu Sun.

PLoS Computational Biology (PCBI), 2025.

 

 

Multimodal Learning

​

• Orthogonalized kernel debiased machine learning for multimodal data analysis. [journal][reprint][preprint]

Xiaowu Dai and Lexin Li.

Journal of the American Statistical Association (JASA), 2022.​

​​

• On MM algorithms for decentralized data. [pdf][code]

Yudong Wang, Jie Hu, Xiaowu Dai, Raymond Carroll, and Yong Chen.

Preprint, 2026.

​​

• Uncertainty-aware multimodal learning via conformal Shapley intervals. [pdf​][code]

Mathew Chandy, Michael Johnson, Judong Shen, Devan V. Mehrotra, Hua Zhou, Jin Zhou, and Xiaowu Dai*.

Preprint, 2026.

​

 

Kernel-Based Learning​

​

• Kernel knockoffs selection for nonparametric additive models. [journal][reprint][preprint][code]

Xiaowu Dai*, Xiang Lyu, and Lexin Li.

Journal of the American Statistical Association (JASA), 2023.

​

• Multi-layer kernel machines: Fast and optimal nonparametric regression with uncertainty quantification. [pdf][code][PyPI]

Xiaowu Dai* and Huiying Zhong.  

Under R&R at Journal of Machine Learning Research (JMLR), 2024.

​

• Effect decomposition of functional-output computer experiments via orthogonal additive Gaussian processes. [pdf]

Yu Tan, Yongxiang Li, Xiaowu Dai, and Kwok-Leung Tsui.

Under R&R at Journal of the American Statistical Association (JASA), 2025.

​

• Variance reduction via resampling and experience replay. [proceedings][reprint][preprint][code]

Jiale Han, Xiaowu Dai*, and Yuhua Zhu.

Fortieth AAAI Conference on Artificial Intelligence (AAAI), 2026.​ Selected for oral presentation.