Nghia Van Vo
Ph.D. student in Applied Mathematics, Oakland University
Advised by Professor Tran T. A. Nghia
I design fast Newton-type algorithms for nonsmooth optimization, and I study when solutions of regularized inverse problems can be trusted under noise. My tools come from variational analysis and generalized differentiation.
I expect to complete my Ph.D. in May 2027 and am applying for tenure-track and postdoctoral positions in mathematics.
Research
Read the research overview-
Fast Newton-type methods for nonsmooth optimization
Second-order algorithms that keep Newton's fast convergence when the objective has a nonsmooth regularizer, such as Lasso, total variation, or OSCAR.
-
Stability of regularized inverse problems
When is a signal recovered from noisy measurements close to the truth? Exact second-order conditions for stable recovery and isolated calmness.
-
Nonconvex optimization in PDE control and machine learning
Inexact DC algorithms in function spaces for sparse optimal control, and per-sample temperatures for knowledge distillation.
Before optimization, I worked on regularity theory for p-Laplace equations (two papers, 2021 and 2022).
Selected publications
All 11 papers-
Nonsmooth Newton methods with effective subspaces for polyhedral regularization
Mathematics of Operations Research, to appear (accepted September 2026)
-
Stable recovery of regularized linear inverse problems
Inverse Problems 41(6), 065018 (2025)
-
Inexact DC algorithms in Hilbert spaces with applications to PDE-constrained optimization
Submitted to Applied Mathematics & Optimization (2026)
-
Consistently informative soft-label temperature for knowledge distillation
Submitted to the International Conference on Artificial Intelligence and Statistics (AISTATS 2027)
News
- Talk at the SIAM DC–Maryland–Virginia Section Meeting, George Mason University.
- Nonsmooth Newton methods with effective subspaces accepted to Mathematics of Operations Research.
- New preprint on isolated calmness of regularized inverse problems.
- New preprint on soft-label temperatures for knowledge distillation.
- Awarded the McKay Outstanding Graduate Research Fellowship, Oakland University.
- Stable recovery of regularized linear inverse problems published in Inverse Problems.
Teaching
Teaching and outreachI lead workshop sections of Elementary Algebra, Intermediate Algebra, and Precalculus at Oakland University, tutor at the Math Help Center, and have guest-taught in the graduate Nonlinear Optimization course. I have also taught in outreach programs for teachers in Vietnam and for middle school students in Detroit.
“The instructor explains the work and will keep asking if you understood it properly.”




