Tam Le - Publications

(*: equal contribution; ⊛: co-last author)

Selected Journals / Conference papers

• Generalized Sobolev Transport for Probability Measures on a Graph
  International Conference on Machine Learning (ICML), 2024. (Acceptance rate: 2609/9473)
  Tam Le*, Truyen Nguyen*, Kenji Fukumizu.

• Sliced Wasserstein with Random-Path Projecting Directions
  International Conference on Machine Learning (ICML), 2024. (Acceptance rate: 2609/9473)
  Khai Nguyen, Shujian Zhang, Tam Le, Nhat Ho.

• Optimal Transport for Measures with Noisy Tree Metric
  International Conference on Artificial Intelligence and Statistics (AISTATS), 2024. (Acceptance rate: 546/1980)
  Tam Le, Truyen Nguyen, Kenji Fukumizu.
  (Preliminary workshop version in NeurIPS workshop: OTML, 2023)

• Scalable Unbalanced Sobolev Transport for Measures on a Graph .
  International Conference on Artificial Intelligence and Statistics (AISTATS), 2023. (Acceptance rate: 496/1686)
  Tam Le*, Truyen Nguyen*, Kenji Fukumizu.

• Sobolev Transport: A Scalable Metric for Probability Measures with Graph Metrics . [Video/Supplementary/POSTER]
  International Conference on Artificial Intelligence and Statistics (AISTATS), 2022. (Acceptance rate: 492/1685)
  Tam Le*, Truyen Nguyen*, Dinh Phung, Viet Anh Nguyen.

• Adversarial Regression with Doubly Non-negative Weighting Matrices. [Supplementary/SLIDE/POSTER]
  Neural Information Processing Systems (NeurIPS/NIPS), 2021. (Acceptance rate: 2334/9122)
  Tam Le*, Truyen Nguyen*, Makoto Yamada, Jose Blanchet, Viet Anh Nguyen.

• Optimal Transport Kernels for Sequential and Parallel Neural Architecture Search. [Supplementary/POSTER]
  International Conference on Machine Learning (ICML), 2021. (Acceptance rate: 1184/5513)
  Vu Nguyen*, Tam Le*, Makoto Yamada, Michael A. Osborne.

• Entropy Partial Transport with Tree Metrics: Theory and Practice. [Video/Supplementary/SLIDE/POSTER]
  International Conference on Artificial Intelligence and Statistics (AISTATS), 2021. (Oral rate: 48/1527=3.1%)
  Tam Le*, Truyen Nguyen*
  (This paper was selected for oral presentation. Oral : acceptance : submissions = 48 : 455 : 1527)

• Flow-based Alignment Approaches for Probability Measures in Different Spaces . [Video/Supplementary/POSTER]
  International Conference on Artificial Intelligence and Statistics (AISTATS), 2021. (Acceptance rate: 455/1527)
  Tam Le*, Nhat Ho*, Makoto Yamada.

• Tree-Sliced Variants of Wasserstein Distances. [Supplementary/POSTER]
  Neural Information Processing Systems (NeurIPS/NIPS), 2019. (Acceptance rate: 1428/6743)
  Tam Le, Makoto Yamada, Kenji Fukumizu, Marco Cuturi.

• Safe Grid Search with Optimal Complexity. [Supplementary/POSTER]
  International Conference on Machine Learning (ICML), 2019. (Acceptance rate: 773/3424)
  Eugene Ndiaye, Tam Le, Olivier Fercoq, Joseph Salmon, Ichiro Takeuchi.

• Persistence Fisher Kernel: A Riemannian Manifold Kernel for Persistence Diagrams. [Supplementary/POSTER]
  Neural Information Processing Systems (NeurIPS/NIPS), 2018. (Acceptance rate: 1011/4856)
  Tam Le, Makoto Yamada.
  (Preliminary workshop version in ICML workshop: GML, 2018)

• Unsupervised Riemannian Metric Learning for Histograms Using Aitchison Transformations. [Video/Supplemental/SLIDE/POSTER]
  International Conference on Machine Learning (ICML), 2015. (Acceptance rate: 270/1037)
  Tam Le, Marco Cuturi.

• Adaptive Euclidean Maps for Histograms: Generalized Aitchison Embeddings
  Machine Learning Journal (MLJ), 2014.
  Tam Le, Marco Cuturi.
  (Preliminary conference version in ACML, 2013)

Others

Other Journals / Conference papers / Preprints

• Tree-Sliced Wasserstein Distance on a System of Lines
  ArXiv, 2024.
  Viet-Hoang Tran*, Trang Pham*, Tho Tran, Tam Le⊛, Tan M. Nguyen⊛.

• SAVA: Scalable Learning-Agnostic Data Valuation
  ArXiv, 2024.
  Samuel Kessler*, Tam Le*, Vu Nguyen.

• Spherical Tree-Sliced Wasserstein Distance
  Under review, 2024.
  Viet-Hoang Tran*, Thanh Chu*, Khoi Nguyen, Trang Pham, Tam Le⊛, Tan M. Nguyen⊛.

• Distance-Based Tree-Sliced Wasserstein Distance
  Under review, 2024.
  Viet-Hoang Tran*, Khoi Nguyen*, Trang Pham, Thanh Chu, Tam Le⊛, Tan M. Nguyen⊛.

• Scalable Counterfactual Distribution Estimation in Multivariate Causal Models
  Conference on Causal Learning and Reasoning (CLeaR), 2024.
  Thong Pham, Shohei Shimizu, Hideitsu Hino, Tam Le.

• Dynamic Flows on Curved Space Generated by Labeled Data
  International Joint Conference on Artificial Intelligence (IJCAI), 2023. (Acceptance rate: 15% out of 4566 submissions)
  Xinru Hua, Truyen Nguyen, Tam Le, Jose Blanchet, Viet Anh Nguyen

• Few-Shot Object Detection via Synthetic Features with Optimal Transport
  ArXiv, 2023.
  Anh Khoa Nguyen Vu, Thanh-Toan Do, Vinh-Tiep Nguyen, Tam Le, Minh-Triet Tran, Tam V. Nguyen

• Point-set Distances for Learning Representations of 3D Point Clouds
  International Conference on Computer Vision (ICCV), 2021. (Acceptance rate: 1617/6236)
  Trung Nguyen, Quang-Hieu Pham, Tam Le, Tung Pham, Nhat Ho, Binh-Son Hua.

• LSMI-Sinkhorn: Semi-supervised Squared-Loss Mutual Information Estimation with Optimal Transport
  European Conference on Machine Learning and Principles & Practice of Knowledge Discovery in Databases (ECML-PKDD), 2021. (Acceptance rate: 147/685)
  Yanbin Liu*, Makoto Yamada*, Yao-Hung Hubert Tsai, Tam Le, Ruslan Salakhutdinov, Yi Yang

• Topological Bayesian Optimization with Persistence Diagrams
  European Conference on Artificial Intelligence (ECAI), 2020. (Acceptance rate: 365/1363)
  Tatsuya Shiraishi, Tam Le, Hisashi Kashima, Makoto Yamada

• Tree-Wasserstein Barycenter for Large-Scale Multilevel Clustering and Scalable Bayes
  ArXiv, 2019.
  Tam Le*, Viet Huynh*, Nhat Ho*, Dinh Phung, Makoto Yamada

• A General Representation Scheme for Crystalline Solids based on Voronoi-Tessellation Real Feature Values and Atomic Property Data
  Science and Technology of Advance Materials Journal (STAM), 2018.
  Randy Jalem, Masanobu Nakayama, Yusuke Noda, Tam Le, Ichiro Takeuchi, Yoshitaka Tateyama, Hisatsugu Yamazaki

• Generalized Aitchison Embeddings for Histograms. [SLIDE]
  Asian Conference on Machine Learning (ACML), 2013 (oral). (Oral paper rate: 9.7%)
  Tam Le, Marco Cuturi

• Image Categorization Using Hierarchical Spatial Matching Kernel
  Journal of the Institute of Image Electronics Engineers of Japan (IIEEJ), Vol. 42, No. 2, 2013.
  Tam Le, Yousun Kang, Akihiro Sugimoto
  (Preliminary conference version in ICIAR, 2011)

• Hierarchical Spatial Matching Kernel for Image Categorization. [SLIDE]
  International Conference on Image Processing and Recognition (ICIAR), 2011 (oral). (Oral paper rate: 25%)
  Tam Le, Yousun Kang, Akihiro Sugimoto, Son Tran, Thuc Nguyen

• Realtime Traffic Sign Detection Using Color and Shape-Based Features
  Asian Conference on Intelligent Information and Database Systems (ACIIDS), 2010 (oral).
  Tam Le, Son Tran, Seichii Mita, Thuc Nguyen

Thesis

• Geometry-Aware Learning Algorithms for Histogram Data Using Adaptive Metric Embeddings and Kernel Functions
  PhD thesis, Kyoto University, Japan, 2015.

Workshop Posters / Talks

Workshop Posters

• 2024, Optimal Transport for Measures with Noisy Tree Metric, ISM Open House, Japan.

• 2024, Scalable Robust Optimal Transport for Measures with Noisy Tree Metric, Deep Learning: Theory, Applications, and Implications workshop, Japan.

• 2023, Optimal Transport for Measures with Noisy Tree Metric, NeurIPS workshop: Optimal Transport and Machine Learning (OTML), US.

• 2023, Optimal Transport for Measures on a Graph, ISM Open House, Virtual.

• 2022, Sobolev Transport: A Scalable Metric for Probability Measures with Graph Metrics, ROIS Crosstalk, Virtual.

• Eugene Ndiaye, Tam Le, Olivier Fercoq, Joseph Salmon, Ichiro Takeuchi, Safe Grid Search with Optimal Complexity. The ACML’2019 RIKEN AIP Workshop, Japan, 2019.

• 2019, Tree-Sliced Approximation of Wasserstein Distances, Korea-Japan Machine Learning workshop, Korea. [Poster]

• 2019, Persistence Fisher Kernel: A Riemannian Manifold Kernel for Persistence Diagrams, Korea-Japan Machine Learning workshop, Korea.

• 2018, Persistence Fisher Kernel: A Riemannian Manifold Kernel for Persistence Diagrams, ICML workshop: Geometry in Machine Learning, Sweden.

• 2018, Riemannian Manifold Kernel for Persistence Diagrams, Transatlantic and Transpacific Workshop on Machine Learning and Discrete Optimization, US.

• Eugene Ndiaye, Tam Le, Olivier Fercoq, Joseph Salmon, Ichiro Takeuchi, Optimal Approximation for regularization and validation path. Journees SMAI MODE, France, 2018.

• Eugene Ndiaye, Tam Le, Olivier Fercoq, Joseph Salmon, Ichiro Takeuchi, GridSearch: Complexite et Garantie pour la Validation (GridSearch: Complexity and Guarantee for Validation). Conference Sur L'Apprentissage Automatique (CAp), France, 2018.

Invited Talks

• 2024, Recent Results on Optimal Transport and Its Applications, Statistical Mathematics Seminar Series, Virtual. [SLIDE]

• 2024, Scalable Robust Optimal Transport for Measures with Noisy Tree Metric , Data Science Seminar (Data Descriptive Science), Kyoto, Japan (hybrid).

• 2024, Local Structures for Large-Scale Optimal Transport, The Mathematics of Data program – Workshop on Optimal Transport and PDEs, IMS, NUS, Singapore. [SLIDE]

• 2023, Sobolev Transport, Data Science Seminar (Data Descriptive Science), Tokyo, Japan. [SLIDE]

• 2023, Optimal transport for measures on a graph, Statistical Mathematics Seminar Series, Virtual. [SLIDE]

• 2023, Optimal Transport with Local Structures for Large-Scale Applications, CREST Seminar: Innovation of Deep Structured Models with Representation of Mathematical Intelligence, Virtual. [SLIDE]

• 2022, Optimal Transport and Its Applications on Machine Learning, Statistical Mathematics Seminar Series, Virtual. [SLIDE]

• 2022, Optimal Transport and Its Applications on Machine Learning, Optimal Transport Seminar (Data Descriptive Science), Virtual.

• 2022, Sobolev Transport: A Scalable Metric for Probability Measures on Graphs, Kyoto Machine Learning workshop, Kyoto University, Japan (hybrid).

• 2022, Geometric Approaches for Persistence Diagrams in Topological Data Analysis, Asia Pacific Seminar on Applied Topology and Geometry (APATG), Virtual. [SLIDE]

• 2022, On Scalability of Optimal Transport with Tree/Graph metric, AAAI workshop (OT-SDM: International Workshop on Optimal Transport and Structured Data Modeling), Virtual. [SLIDE]

• 2021, Adversarial Regression with Doubly Non-negative Weighting Matrices, VinAI workshop, Virtual. [SLIDE]

• 2021, An Introduction on Tree-(Sliced)-Wasserstein Geometry, RIKEN AIP Open Seminar Series, Virtual. [SLIDE]

• 2019, Kernel methods for Persistence Diagrams: A Geometric Approach, Institute for Advanced Study, Kyoto University, Japan. [SLIDE]

• 2016, Geometry-Aware Metric Learning for Histograms, MI2I Open Seminar Series, National Institute for Materials Science, Tsukuba, Japan. [SLIDE]

• 2015, Metric Learning for Histograms, Nagoya Institute of Technology, Japan.

• 2015, Unsupervised Riemannian Metric Learning for Histograms, Bioinformatics Center, Institute for Chemical Research, Kyoto University, Japan.

• 2015, Unsupervised Riemannian Metric Learning for Histograms Using Aitchison Embeddings, IST Seminar, Graduate School of Informatics, Kyoto University, Japan. [SLIDE]

• 2014, Adaptive Euclidean Maps for Histograms: Generalized Aitchison Embeddings, Lear, INRIA Rhone Alpes, Grenoble, France. [SLIDE]