My research activities are hereby presented sometimes redundantly, topic by topic, to give different reading keys to my works.
Optimal Transport for Learning on Graphs
How to solve a supervised graph prediction problem leveraging (Fused) Gromov-Wasserstein distances and variants ?
- Melo, G., de Saivre, T., Calissano, A., & d’Alché-Buc, F. (2026). Conformal Graph Prediction with Z-Gromov Wasserstein Distances. arXiv preprint: https://arxiv.org/pdf/2603.02460, to be presented at UAI 2026.
- Paul Krzakala, Gabriel Silva de Melo, Charlotte Laclau, Florence d’Alché-Buc, Remi Flamary. The quest for the GRAph Level autoEncoder (GRALE), arxiv, Advances in Neural Information Processing, NeurIPS 2025.
- Junjie Yang, Matthieu Labeau, Florence d’Alché-Buc: Exploiting Edge Features in Graph-based Learning with Fused Network Gromov-Wasserstein Distance. Trans. Mach. Learn. Res. 2024 (2024)
- Paul Krzakala, Junjie Yang, Rémi Flamary, Florence d’Alché-Buc, Charlotte laclau, Mathieu Labeau. Any2Graph: Deep End-To-End Supervised Graph Prediction With An Optimal Transport Loss, NeurIPS 2024, arxiv here. (2024)
- Luc Brogat-Motte, Rémi Flamary, Céline Brouard, Juho Rousu, Florence d’Alché-Buc: Learning to Predict Graphs with Fused Gromov-Wasserstein Barycenters. Int. Conf. on Machine Learning (ICML), PMLR, 2022. (2022)
Metabolite identification
How to achieve Metabolite Retrieval from Mass Spectra ?
- Céline Brouard, Huibin Shen, Kai Dührkop, Florence d’Alché-Buc, Sebastian Böcker, Juho Rousu. Fast metabolite identification with Input Output Kernel Regression, Bioinformatics, May, 2016.
- Céline Brouard, Antoine Bassé, Florence d’Alché-Buc and Juho Rousu, Improved Small Molecule Identification through Learning Combinations of Kernel Regression Models, Metabolites 2019, 9(8), 160; https://doi.org/10.3390/metabo9080160 – 01 Aug 2019.
- Luc Brogat-Motte, Alessandro Rudi, Céline Brouard, Juho Rousu, Florence d’Alché-Buc, Vector-Valued Least-Squares Regression under Output Regularity Assumptions, Journal of Mach. Learn. Res., 23(344):1−50, 2022.
- Luc Brogat-Motte, Rémi Flamary, Céline Brouard, Juho Rousu, Florence d’Alché-Buc: Learning to Predict Graphs with Fused Gromov-Wasserstein Barycenters. Int. Conf. on Machine Learning (ICML), PMLR, 2022. (2022)
- Junjie Yang, Matthieu Labeau, Florence d’Alché-Buc: Exploiting Edge Features in Graph-based Learning with Fused Network Gromov-Wasserstein Distance. Trans. Mach. Learn. Res. 2024 (2024)
- Krzakala, P., Melo, G., Lançon, C., Laclau, C., Flamary, R., Thévenot, E., & d’Alché-Buc, F. (2026). MSAlign: Aligning Molecule and Mass Spectra Foundation Models for Metabolite Identification. arXiv preprint arXiv:2605.19752.
Low-rank and sketching approaches for Frugality
How to apply low-rank approaches (reduced rank, sketching) in vector-valued kernel methods to get the best efficiency both in terms of statistics and space/compute cost ?
- Brogat-Motte, L., Rudi, A., Brouard, C., Rousu, J., & d’Alché-Buc, F. (2022). Vector-valued least-squares regression under output regularity assumptions. Journal of Machine Learning Research, 23(344), 1-50. article
- El Ahmad, T. , Laforgue, P., & d’Alché-Buc, F. (2022). Fast Kernel Methods for Generic Lipschitz Losses via $ p $-Sparsified Sketches, Trans. Mach. Learn. Res. 2023 (2023) arXiv preprint arXiv:2206.03827.
- El Ahmad, T., Brogat-Motte, L., Laforgue, P., & d’Alché-Buc, F. (2024, April). Sketch in, sketch out: Accelerating both learning and inference for structured prediction with kernels. In International conference on artificial intelligence and statistics (pp. 109-117). PMLR.Link
How to make deep neural network work in a RKHS subspace and target structured prediction ?
- El Ahmad, T., Yang, J., Laforgue, P., & d’Alché-Buc, F. (2024, August). Deep sketched output kernel regression for structured prediction. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases (pp. 93-110). arxiv