Anna Ma
University of California-Irvine, CA·Mathematics & Statistics
Much real-world data comes as tensors, which are multi-dimensional arrays such as a table that also changes over time. These datasets are often incomplete, noisy or only partly observed, and that limits how reliable data-driven systems and AI can be.
- Fields
Mathematics & Statistics
Computer Science
- Topics
tensor completion
incomplete tensors
low-rank tensors
- Grants
- 1 known grant, through Aug 2029
What they work on
- Analyze the sample complexity of low-rank tensor completion, meaning how many observed entries are needed to recover a multi-dimensional dataset with simple underlying structure.
- Develop a randomized sampling scheme for tensor completion and show it allows polynomial-time recovery of low-rank tensors with nearly linear numbers of samples.
- Design and analyze numerical algorithms for tensor completion when missing entries follow a structured pattern rather than being randomly scattered.
- Develop and analyze completion-free optimization methods that work directly with incomplete tensors when filling in the missing data is not feasible.
More about the research
This project builds theory and algorithms for working with incomplete tensors. It asks two questions: how much data is needed to fill in the missing entries, and when filling them in is not possible, how to compute useful answers anyway.
Why it matters
Many AI and data analytics systems depend on datasets that are partly missing or noisy. Methods that recover structure from far fewer observations, or that work without filling in gaps at all, could help companies and scientists build more accurate, efficient and trustworthy data-driven tools.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
HanQin Cai
University of Central Florida, FL·Mathematics & Statistics
Develop scalable tensor algorithms based on mode-wise sampling and low-rank tensor representations to recover useful information from large, partially corrupted or incomplete multidimensional data.
Longxiu Huang
Michigan State University, MI·Mathematics & Statistics
Develop mode-wise sampling methods that analyze large tensors (multidimensional data arrays) by working with small slices along each dimension, and prove that this sampling reliably captures the full data.
Fabian Faulstich
Rensselaer Polytechnic Institute, NY·Mathematics & Statistics·Early-career
Recast periodic coupled-cluster theory as a nonlinear Petrov–Galerkin method for a pseudo-differential operator on a high-dimensional torus, in order to derive sharp error estimates for electronic-structure simulations of crystalline materials.
The school
University of California-Irvine: 46 labs on PI Pages, $645M a year on research, 524 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.