Hedi Xia

Contact

• Email: hedixia@math.ucla.edu
• Office: MS 5343
• Github: https://github.com/hedixia

About

I am currently a third year Ph.D. students in Applied Mathematics at UCLA and I am working under the supervision of Stanley Osher and Andrea Bertozzi. I did my undergraduate in Mathematics at College of Creative Studies, UC Santa Barbara with a minor in Statistical Science in 2015 to 2019.

Research Interest

• Machine Learning
• Time Series
• Graph Neural Networks

Conference & Publications

• Heavy Ball Neural Ordinary Differential Equations
  • Hedi Xia$^1$, Vai Suliafu$^1$, Tan Minh Nguyen, Hangjie Ji, Andrea Bertozzi, Stanley Osher, and Bao Wang
  • Neural Information Processing Systems, 2021.
  • [OpenReview], [Arxiv], [Github]
• Kernel Treelets
  • Hedi Xia and Hector D. Ceniceros
  • Advances in Data Science and Adaptive Analysis, 2019. doi:10.1142/S2424922X19500062.
  • [ADSAA], [Arxiv], [Github]

Preprints

• How Does Momentum Benefit Deep Neural Networks Architecture Design? A Few Case Studies
  • Bao Wang, Hedi Xia, Tan Nguyen, Stanley Osher
  • [Arxiv]

Teaching

• PIC 16A Python with Applications (x2), Spring 2021
• Math 164 Optimization, Fall 2020
• Math 142 Mathematical Modeling, Fall 2020
• Math 142 Mathematical Modeling, Spring 2020
• PIC 16 Python with Applications, Winter 2020

[1] Cofirst Author