Jianhao Ma
 |
Jianhao Ma |
About me
I am currently a postdoctoral researcher at the University of Pennsylvania, working with Professor Yuxin Chen. I received my Ph.D. in Operations Research from the University of Michigan, where I was fortunate to be advised by Professor Salar Fattahi. My research focus on developing efficient and robust machine learning models. I welcome future collaborations. Please feel free to contact me via Email!
Research Interests
General optimization and generalization in modern machine learning.
Structured nonsmooth optimization.
Quantization and pruning.
Preprints
Preconditioning Benefits of Spectral Orthogonalization in Muon [arxiv]
Jianhao Ma, Yu Huang, Yuejie Chi, Yuxin Chen
Certifying Optimality in Nonconvex Robust PCA [arxiv]
Pinxi Gong, Lexiao Lai, Jianhao Ma
Quantization through Piecewise-Affine Regularization: Optimization and Statistical Guarantees [arxiv]
Jianhao Ma, Lin Xiao
Implicit Regularization of Infinitesimally-perturbed Gradient Descent Toward Low-dimensional Solutions [arxiv]
Jianhao Ma, Geyu Liang, Salar Fattahi
RANSAC Revisited: An Improved Algorithm for Robust Subspace Recovery under Adversarial and Noisy Corruptions [arxiv]
Guixian Chen, Jianhao Ma, Salar Fattahi
LIBRA: Language Model Informed Bandit Recourse Algorithm for Personalized Treatment Planning [SSRN]
Junyu Cao, Ruijiang Gao, Esmaeil Keyvanshokooh, Jianhao Ma
Publications
Can Learning Be Explained By Local Optimality In Low-rank Matrix Recovery? [paper][arxiv]
Mathematics of Operations Research 2025
INFORMS Junior Faculty Interest Group (JFIG) Paper Competition – Second Place (as a student coauthor)
Jianhao Ma, Salar Fattahi
PARQ: Piecewise-Affine Regularized Quantization [paper]
International Conference on Machine Learning (ICML) 2025
Lisa Jin, Jianhao Ma, Zechun Liu, Andrey Gromov, Aaron Defazio, Lin Xiao
Convergence of Gradient Descent with Small Initialization for Unregularized Matrix Completion [paper][arxiv]
Conference on Learning Theory (COLT) 2024
Jianhao Ma, Salar Fattahi
Robust Sparse Mean Estimation via Incremental Learning [arxiv]
ICLR Workshop on Bridging the Gap Between Practice and Theory in Deep Learning 2024
Jianhao Ma, Rui Ray Chen, Yinghui He, Salar Fattahi, Wei Hu
Global Convergence of Sub-gradient Method for Robust Matrix Recovery: Small Initialization, Noisy Measurements, and Over-parameterization [paper][arxiv]
Journal of Machine Learning Research (JMLR) 2023
Jianhao Ma, Salar Fattahi
Behind the Scenes of Gradient Descent: A Trajectory Analysis via Basis Function Decomposition [paper][arxiv]
International Conference on Learning Representations (ICLR) 2023
Jianhao Ma, Lingjun Guo, Salar Fattahi
Blessing of Nonconvexity in Deep Linear Models: Depth Flattens the Optimization Landscape Around the True Solution [paper][arxiv]
Advances in Neural Information Processing Systems (NeurIPS) 2022 (Spotlight)
Katta Murty Prize for Best Research Paper on Optimization
Jianhao Ma, Salar Fattahi
Towards Understanding Generalization via Decomposing Excess Risk Dynamics [paper] [arxiv]
International Conference on Learning Representations (ICLR) 2022
Jiaye Teng*, Jianhao Ma*, Yang Yuan
Sign-RIP: A Robust Restricted Isometry Property for Low-rank Matrix Recovery [paper][arxiv]
Jianhao Ma, Salar Fattahi
NeurIPS Workshop on Optimization for Machine Learning, 2021
News
October 2024: I will attend INFORMS Annual Meeting.
March 2024: I will attend INFORMS Optimization Society Conference.
March 2024: Our paper has been accepted at ICLR BGPT workshop.
March 2024: Thrilled to receive Rackham Predoctoral Fellowship.
February 2024: I will intern at FAIR in Meta with Dr. Lin Xiao this summer.
February 2024: New paper about the global convergence of GD for unregularized matrix completion.
Services
Reviewer of SIOPT, IEEE TIT, IEEE TSP, JMLR, ICML, NeurIPS, ICLR, AISTATS.
Quate
Always consider a problem under the minimum structure in which it makes sense. —— Gustave Choquet