Education
- Ph.D. in Applied Mathematics, University of California, Berkeley, Sep. 2023 -
- M.S. in Computational and Applied Mathematics, The University of Chicago, Sep. 2021 - Dec. 2022
- B.S. in Statistics, Xiamen University, Sep. 2017 - Jun. 2021
Research Papers
QEDBench: Quantifying the Alignment Gap in Automated Evaluation of University-Level Mathematical Proofs
Santiago Gonzalez, et al. (including Yuchen Fang)
Submitted
3DGS$^2$-TR: A Scalable Second-Order Trust-Region Method for 3D Gaussian Splatting
Roger Hsiao, Yuchen Fang, Xiangru Huang, Ruilong Li, Hesam Rabeti, Zan Gojcic, Javad Lavaei, James Demmel,and Sophia Shao
Submitted
Why is Normalization Preferred? A Worst-Case Complexity Theory for Stochastically Preconditioned SGD under Heavy-Tailed Noise
Yuchen Fang, James Demmel, and Javad Lavaei
Submitted
TRSVR: An Adaptive Stochastic Trust-Region Method with Variance Reduction
Yuchen Fang, Xinshou Zheng, and Javad Lavaei
Submitted
On the Sharp Input-Output Analysis of Nonlinear Systems under Adversarial Attacks
Jihun Kim, Yuchen Fang, and Javad Lavaei
Submitted
High-Probability Complexity Bounds of Trust-Region Stochastic Sequential Quadratic Programming with Heavy-Tailed Noise
Yuchen Fang, Javad Lavaei, and Sen Na
Submitted
Trust-Region Sequential Quadratic Programming for Stochastic Optimization with Random Models
Yuchen Fang, Sen Na, Michael W. Mahoney and Mladen Kolar
Submitted
Fully Stochastic Trust-Region Sequential Quadratic Programming for Equality-Constrained Optimization Problems
Yuchen Fang, Sen Na, Michael W. Mahoney and Mladen Kolar
SIAM Journal on Optimization, 2024
Workshop Papers
Trust-Region Sequential Quadratic Programming for Equality-Constrained Stochastic Optimization with Random Models: First-Order Stationarity (workshop version)
Yuchen Fang, Sen Na and Mladen Kolar
Higher-Order Optimization in Machine Learning (HOOML) 2022, NeurIPS Workshop
Fully Stochastic Trust-Region Sequential Quadratic Programming for Equality-Constrained Optimization Problems (workshop version)
Yuchen Fang, Sen Na and Mladen Kolar
Higher-Order Optimization in Machine Learning (HOOML) 2022, NeurIPS Workshop
Presentations
Dec. 2022, “Trust-Region Sequential Quadratic Programming for Stochastic Optimization with Random Models: First-Order Stationarity”, NeurIPS Workshop on Higher-Order Optimization in Machine Learning (HOOML) 2022
Dec. 2022, “Fully Stochastic Trust-Region Sequential Quadratic Programming for Equality-Constrained Optimization Problems”, NeurIPS Workshop on Higher-Order Optimization in Machine Learning (HOOML) 2022
Nov. 2022, “Fully Stochastic Trust-Region Sequential Quadratic Programming for Equality-Constrained Optimization Problems”, Computational and Applied Mathematics (CAM) Student Seminar, the University of Chicago
May 2022, “Trust-Region Stochastic Sequential Quadratic Programming for Constrained Optimization”, ACMNTW Workshop on Optimization and Machine Learning 2022, Northwestern University
Professional Service
- Conference Reviewer:
- International Conference on Learning Representations (ICLR) 2024,2025
- Conference on Neural Information Processing Systems (NeurIPS) 2023
- International Conference on Artificial Intelligence and Statistics (AISTATS) 2023
- Midwest Machine Learning Symposium 2023
- Teaching:
- GSI of Math 1A Calculus, University of California, Berkeley, Fall 2023 - Fall 2024
- Grader of STAT 37710, the University of Chicago, Autumn 2022
Skills
- Programming Languages: R, Python, MATLAB, Julia, C, LaTeX
- Languages: Native in Mandarin, fluent in English