About me
Hoang-Son TRAN
PhD Candidate
Department of Mathematics
National University of Singapore
Email address 1: hoangson.tran at nus.edu.sg
Email address 2: transonsp97 at gmail.com
Welcome to my homepage. My name is Hoang-Son TRAN (Trần Hoàng Sơn in Vietnamese). I am a PhD student in Mathematics at National University of Singapore (NUS).
Useful links: Google Scholar , CV
Education
August 2020 to 2026: PhD candidate, Department of Mathematics, National University of Singapore, Singapore
Advisors: Professor Subhro GHOSH and Professor Tien-Cuong DINH
Thesis: Strongly correlated interacting particle systems and their applications.
2015-2019: Bachelor of Mathematics, Hanoi National University of Education, Hanoi, Vietnam
2012-2015: High School for Gifted Students, Hanoi National University of Education, Hanoi, Vietnam
Research
My research develops probabilistic and statistical-mechanical methods for machine learning, with a focus on negatively dependent particle systems, determinantal point processes (DPPs), concentration inequalities, central limit theorems, and variance reduction. A central goal is to build mathematically principled sampling and subsampling methods with provable guarantees for learning and approximation.
State-of-art minibatches via novel DPP kernels: discretization, wavelets, and rough objectives, preprint, arXiv:2605.13127, (with P. Gupta, R. Bardenet, and S. Ghosh), 2026
Gaussian mixtures and non-parametric likelihoods through the lens of statistical mechanics, preprint, arXiv:2603.23196, (with S. Ghosh, A. Guntuboyina, and S. Mukherjee), 2026
Negative dependence as a toolbox for machine learning: review and new developments, preprint, arXiv:2502.07285 , (with R. Bardenet, S. Ghosh, and V. Petrovic), 2025
Small coresets via negative dependence: DPPs, linear statistics, and concentration, NeurIPS spotlight (< 2.08% total submissions), NeurIPS Proceedings , (with R. Bardenet, S. Ghosh, and H. Simon), 2024
Learning Networks from Gaussian Graphical Models and Gaussian Free Fields, Journal of Statistical Physics, rdcu.be/dDio3 , (with S. Ghosh, S. S. Mukherjee, and U. Gangopadhyay), 2024
Gaussian fluctuations for spin systems and point processes: near-optimal rates via quantitative Marcinkiewicz's theorem, minor revision at Annals of Applied Probability, arxiv.org/abs/2107.08469, (with T.-C. Dinh, S. Ghosh, and M.-H. Tran), 2021
Awards
Best Graduate Researcher Award, Department of Mathematics, NUS, 2025
Faculty of Science PhD Conference Award, NUS, 2024
Graduate Tutor Merit Award, Department of Mathematics, NUS, 2024
Graduate Tutor Commendation Letters, Department of Mathematics, NUS, 2023
First Prize in Vietnam Mathematical Olympiad for University Students, 2017
First Prize in Vietnam Mathematical Olympiad for University Students, 2016
Second Prize in Vietnam Mathematical Olympiad, 2015
Seminars and Activities
I am co-organizing:
Reading group on "Statistical Optimal Transport" at NUS
Reading group on "Log-concave Sampling" (online)
One day workshop "Mathematics of Machine Intelligence" at NUS, 29 April 2025
Conformal Field Theories: Randomness and Geometry at IMS, 18-29 May 2026
Referee for Electronic Journal of Probability, Biometrika.
Selected Talks
“Central limit theorems for strongly correlated models via quantitative Marcinkiewicz’s theorem”, Stat-Math Seminars and Colloquium, Indian Statistical Institute, Bangalore Centre.
“Determinantal point processes in machine learning: smaller coresets via negative dependence”, 2025 INFORMS International Meeting, Singapore.
“Small coresets via negative dependence: DPPs, linear statistics, and concentration”, NeurIPS 2024, Vancouver, Canada.