I am a third-year PhD student in Computer Science at Purdue University advised by Prof. Steve Hanneke. I am working on foundations of machine learning. My research interests lie broadly at the intersection of learning theory and algorithmic statistics, as well as their potential extensions to understanding modern artificial intelligence. I studied Mathematics at SJTU for undergraduate. I am fortunate enough to be able to work with several erudite professors and fantastic collaborators, including Prof. Daniel Hsu, Prof. Samory Kpotufe, Prof. Itay Safran, Prof. Gal Vardi, Gan Yuan, Hongao Wang.
Research Interests
To understand the statistical limits of the learning performance under variants of the PAC (Probably Approximately Correct) model such as universal learning, multitask learning, meta learning and multi-distribution learning.
To develop efficient learning/estimating algorithms that admit certain beneficial aspects such as sample efficiency, adaptivity, robustness. Related topics include algorithmic data selection, adaptive learning, adversarially robustness, etc.
To provably argue the tremendous success of modern deep learning architectures (e.g., neural networks, transformers), with a focus on their impressive generalization capabilities.
\( \small (\alpha-\beta) \): alphabetical ordering, *: equal contributions
Preprint, 2026
Steve Hanneke, Hongao Wang, Mingyue Xu \( \small (\alpha-\beta) \).
The 43rd International Conference on Machine Learning (ICML, 2026 - Spotlight)
Mingyue Xu, Gal Vardi, Itay Safran.
The 43rd International Conference on Machine Learning (ICML, 2026)
Steve Hanneke, Mingyue Xu \( \small (\alpha-\beta) \).
The 38th Annual Conference on Learning Theory (COLT, 2025)
Steve Hanneke, Mingyue Xu \( \small (\alpha-\beta) \).
SIAM Journal on Mathematics of Data Science (SIMODS, 2025)
Gan Yuan\(^*\), Mingyue Xu\(^*\), Samory Kpotufe, Daniel Hsu.
The 38th Annual Conference on Neural Information Processing Systems (NeurIPS, 2024)
Steve Hanneke, Mingyue Xu \( \small (\alpha-\beta) \).
IEEE Transactions on Information Theory (TIT, 2023)
Jiyuan Tu, Weidong Liu, Xiaojun Mao, Mingyue Xu.
Teaching assistant, Purdue University
CS182: Foundations of Computer Science, Spring 2025, Spring 2026.
CS251: Data Structures and Algorithms, Fall 2023, Spring 2024, Fall 2024.
CS381: Introduction to the Analysis of Algorithms, Fall 2025.
Professional service
Reviewer: International Conference on Learning Representations (ICLR), 2026.
Gold Reviewer: International Conference on Machine Learning (ICML), 2026.
Reviewer: Neural Information Processing Systems (NeurIPS), 2026.
I am always happy to discuss research ideas and potential collaborations.
Feel free to contact me if you are interested in my works.