Mathematical Statistics · AI for Science · Mixture of Experts

TrungTin Nguyen

Vietnamese name: Nguyễn Trung Tín. I am a mathematical statistician working at the interface of statistical theory, machine learning, and AI for science. My research develops principled methods for model selection, uncertainty quantification, and mixture-of-experts learning, with applications in genomics, systems biology, and trustworthy AI.

I have tried to make my research something you can run, not just read: 23 interactive demos turn my papers into experiments you can explore in the browser, and eleven open courses offer lecture notes, 566 teaching demonstrations and Python and R code, from first-year statistics to advanced probabilistic machine learning.

Portrait of TrungTin Nguyen

Profile highlights

Current roleMACSYS Postdoctoral Research Fellow / Lecturer, School of Mathematical Sciences, Queensland University of Technology; member of the QUT Centre for Data Science
Scholarly record40 outputs: 11 refereed journal articles, 15 refereed conference papers, 5 refereed workshop papers, 2 contributed conference talks, 1 PhD thesis and 6 preprints; 646 citations (641 since 2021), h-index 13, i10-index 16 (Google Scholar, 19 November 2026)
Recent highlightsAISTATS 2026 Spotlight, ICLR 2026, two papers in the Annals of the Institute of Statistical Mathematics (2026), Econometrics and Statistics (2026), two NeurIPS 2025 papers, Bayesian Analysis (2025); NeurIPS 2026 Top Reviewer

About

I build rigorous statistical and machine learning methods that combine mathematical foundations with real scientific impact. My recent work focuses on model selection, mixture-of-experts models, uncertainty quantification, and scalable learning algorithms for heterogeneous and high-dimensional data.

Biography

I am currently a Postdoctoral Research Fellow / Lecturer (Level B) in the School of Mathematical Sciences at Queensland University of Technology, Brisbane, Australia. I contribute to the ARC Centre of Excellence for the Mathematical Analysis of Cellular Systems (MACSYS) and the QUT Centre for Data Science, where I work on foundational methods in data science and AI.

My broader goal is to develop mathematically grounded learning methods that remain interpretable, robust, and useful in scientific applications. This includes principled methodology for heterogeneous data, Bayesian inference, uncertainty-aware prediction, and statistical learning for biological systems; in MACSYS, for example, I develop Bayesian parameter balancing, which turns inconsistent literature values of enzyme-kinetic parameters into calibrated uncertainty for kinetic models of cells.

Career

  • 2025 to now: MACSYS Postdoctoral Research Fellow / Lecturer, School of Mathematical Sciences, Queensland University of Technology
  • 2023 to 2024: Postdoctoral Research Fellow, School of Mathematics and Physics, The University of Queensland
  • 2022 to 2023: Postdoctoral Research Fellow, Statify team, Inria Centre at Université Grenoble Alpes, France

Education

  • PhD: Statistics and Data Science, Normandie Université
  • MSc: Applied Mathematics, Université d’Orléans
  • BSc: Probability and Statistics, VNU-HCM University of Science

Research

My research program sits at the intersection of mathematical statistics, machine learning, and scientific applications. I aim to design methods that are theoretically justified, computationally scalable, and practically valuable.

Statistical theory

Model selection, oracle inequalities, penalisation, minimal penalties, missing-data methodology, and non-asymptotic analysis for heterogeneous and high-dimensional models.

Mixture-of-experts models

Approximation theory, parameter estimation, identifiability, scalable optimisation, sparse modelling, and uncertainty-aware inference for mixture-of-experts and related conditional mixture models.

Bayesian computation

Bayesian synthetic likelihood, surrogate posteriors, MCMC, variational methods, uncertainty quantification, and principled probabilistic inference under computational and modelling constraints.

AI for science

Kinetic models of cells (MACSYS), genomics and transcriptomics, medical vision-language models and molecular property prediction, with a focus on interpretable and trustworthy methods for real scientific discovery.

Selected publications

View full publication catalogue

A selection of recent and representative publications across mathematical statistics, mixture-of-experts models, Bayesian computation, and AI for science.

2026 · AISTATS 2026 Spotlight

Dendrograms of Mixing Measures for Softmax-Gated Gaussian Mixture of Experts: Consistency Without Model Sweeps

Do Tien Hai; Trung Nguyen Mai; TrungTin Nguyen; Nhat Ho; Binh T Nguyen; Christopher Drovandi.

2026 · ICLR 2026

FACET: A Fragment-Aware Conformer Ensemble Transformer

Duy M. H. Nguyen; Trung Q. Nguyen; Ha T. H. Le; Mai Thanh Nhat Truong; TrungTin Nguyen; Nhat Ho; Khoa D. Doan; Duy Duong-Tran; Li Shen; Daniel Sonntag; James Zou; Mathias Niepert; Hyojin Kim; Jonathan E. Allen.

2026 · Annals of the Institute of Statistical Mathematics

Non-asymptotic oracle inequalities for the Lasso in high-dimensional mixture of experts

TrungTin Nguyen; Hien D. Nguyen; Faicel Chamroukhi; Geoffrey J. McLachlan.

2026 · bioRxiv preprint

Bayesian Parameter Balancing Enables Robust and Consistent Estimation of Kinetic Parameter Uncertainty

TrungTin Nguyen; Binh H. Ho; Michael Pan; Jennifer A. Flegg; Michael J. McDonald; Christopher Drovandi.

2025 · NeurIPS 2025

A Unified Framework for Variable Selection in Model-Based Clustering with Missing Not at Random

Binh H Ho; Long Nguyen Chi; TrungTin Nguyen; Binh T Nguyen; Van Ha Hoang; Christopher Drovandi.

2025 · NeurIPS 2025

ExGra-Med: Extended Context Graph Alignment for Medical Vision-Language Models

Duy M. H. Nguyen; Nghiem T. Diep; Trung Q. Nguyen; Hoang-Bao Le; Tai Nguyen; Tien Nguyen; TrungTin Nguyen; Nhat Ho; Pengtao Xie; Roger Wattenhofer; James Zou; Daniel Sonntag; Mathias Niepert.

2024 · NeurIPS 2024

Accelerating Transformers with Spectrum-Preserving Token Merging

Hoai-Chau Tran; Duy MH Nguyen; Manh-Duy Nguyen; TrungTin Nguyen; Ngan Hoang Le; Pengtao Xie; Daniel Sonntag; James Zou; Binh T. Nguyen; Mathias Niepert.

2023 · EMNLP 2023 Main

HyperRouter: Towards Efficient Training and Inference of Sparse Mixture of Experts

Truong Giang Do; Khiem Le; Quang Pham; TrungTin Nguyen; Thanh-Nam Doan; Binh T. Nguyen; Chenghao Liu; Savitha Ramasamy; Xiaoli Li; Steven Hoi.

2022 · Electronic Journal of Statistics

A non-asymptotic approach for model selection via penalization in high-dimensional mixture of experts models

TrungTin Nguyen; Hien Duy Nguyen; Faicel Chamroukhi; Florence Forbes.

Website reach

Recent website activity from Google Analytics 4. These figures provide a lightweight view of the website’s academic reach and are updated automatically each week from analytics/summary.json.

Active users at last update–GA4 activeUsers, as of the last weekly refresh
Users in last 30 days–Total users over the recent 30-day window
Sessions in last 30 days–Total visits (sessions) over the recent 30-day window
Page views in last 30 days–Screen/page views over the recent 30-day window
Total users recorded by GA4–Calculated from 1 Jan 2020 to today
Total sessions recorded by GA4–Total visits (sessions) from 1 Jan 2020 to today
Total page views recorded by GA4–Screen/page views from 1 Jan 2020 to today
Top viewer locationsLast 30 days by users
Waiting for GA4 data
Top pagesLast 30 days by page views
Waiting for GA4 data

Last updated: not yet connected