I am currently pursuing my Ph.D. in Statistics at Penn State. This summer, I joined Amazon as an applied scientist intern, working on using time-series foundation models for anomaly detection. I am advised by Dr. Runze Li and in collaboration with Dr. Ying Sun from the EECS department. My current research focuses on designing transfer-learning/meta-learning methods for large-scale dataset modeling under crowd-sourcing settings.

Publications and Preprints

• Weakly-supervised Multi-sensor Anomaly Detection with Time-series Foundation Models (2024+)
  • Zelin He, Matthew Reimherr, Sarah Alnegheimish, Akash Chandrayan
  • NeurIPS 2024 TSALM Workshop
• TransFusion: Covariate-Shift Robust Transfer Learning for High-Dimensional Regression (2024) [Paper] [Code]
  • Zelin He, Ying Sun, Jingyuan Liu, Runze Li
  • Proceedings of The 27th International Conference on Artificial Intelligence and Statistics (AISTATS)
• AdaTrans: Feature-wise and Sample-wise Adaptive Transfer Learning for High-dimensional Regression (2024+) [Paper] [Slides] [Poster] [Talk (in Chinese)]
  • Zelin He, Ying Sun, Jingyuan Liu, Runze Li
• The Effect of Personalization in FedProx: A Fine-grained Analysis on Statistical Accuracy and Communication Efficiency (2024+) [Paper] [Code]
  • Xin Yu*, Zelin He*, Ying Sun, Lingzhou Xue, Runze Li

📧 If you’re interested in my research, don’t hesitate to send me an email!! Let’s chat more about it.

Reviewer Experience

• ICML 2024
• ICLR 2025

Education

• Ph.D. in Statistics, Pennsylvania State University, 2022 - Present.
• B.S. in Statistics, Beijing Normal University, 2018 - 2022.

Awards

• University Graduate Fellowship, 2023
• Early Pass of the Doctoral Theory Qualifying Exam, 2022
• Paul M. Doty Distinguished Graduate Fellowship, 2022
• Verne M. Willaman Distinguished Graduate Fellowships in Science, 2022
• National Scholarship, 2021