About

I am a Postdoctoral Researcher at Telecom Paris, working with Florence d’Alché, Gabriel Peyré and Remi Flamary on the applications of optimal transport to machine learning, in particular to graph-structured data.

I obtained my PhD in Proba-Stat team of Département de Mathématiques d’Orsay and INRIA Parietal team. I was very fortunate to be advised by Sylvain Arlot and Bertrand Thirion. My doctoral thesis dealt with high-dimensional statistics on both theoretical and algorithmic aspects.

I also spent a fair amount of time during my PhD on the developments of open-source softwares, and hope to continue to do so in the future. Checkout Nilearn, Hidimstat, and Benchopt as some of the open-source projects that I have been developing/contributing to.

Contact

  • binguyen [AT] telecom-paris.fr or tuanbinhs [AT] gmail.com (the former redirects to the latter anyway).

Research

(see also my Google Scholar)

  • BN, Thirion, B., & Arlot, S. A Conditional Randomization Test for Sparse Logistic Regression in High-Dimension. NeuRIPS 2022. [paper] [code] [reviews]

  • Moreau, T. et al. Benchopt: Reproducible, efficient and collaborative optimization benchmarks. NeuRIPS 2022. [paper] [code] [reviews]

  • J.-A. Chevalier, BN, B. Thirion J. Salmon, Spatially relaxed inference on high-dimensional linear models. To appear in Journal of Statistics and Computing, 2022. [paper]

  • J.-A. Chevalier, BN, J. Salmon, G. Varoquaux, B. Thirion, Decoding with confidence: Statistical control on decoder maps. In NeuroImage, Volume 234, 2021, 117921, ISSN 1053-8119. [paper]

  • BN, J.-A. Chevalier, B.Thirion, & S. Arlot, Aggregation of Multiple Knockoffs. In Proceedings of the 37th International Conference on Machine Learning (ICML); PMLR 119:7283-7293, 2020. [paper] [code]

  • BN, J.-A Chevalier, & B. Thirion, ECKO: Ensemble of Clustered Knockoffs for Robust Multivariate Inference on fMRI Data. In International Conference on Information Processing in Medical Imaging (pp. 454-466). Springer, Cham., 2019 [paper]

Slides

  • Slides of my thesis defense.