Member of LTCI, Dept Image Data and Signal, S2A Team.
Ellis Fellow. Member of the Ellis Board. In charge of the Ellis PhD programme.
Holder of the Télécom Paris Chair about Data Science and Artificial Intelligence for digitalised industry and services,
Funded by industrial partners: Airbus Defense and Space, ENGIE, IDEMIA, SAFRAN, VALEO – Started in 2019.
Formerly
- Head of the Image, Data and Signal Department at Télécom Paris – september 2021 – Oct 2023.
- head of DigiCosme, the ANR Excellence laboratory in Computer Science of Paris-Saclay University 2017-19
Research Interests
Machine Learning & Artificial Intelligence, bioinformatics & medical applications, industrial applications.
Subdomains: (Operator-valued) Kernel Methods, Structured Output Prediction, Complex data, Reliable Machine Learning, Dynamical Systems Modeling.
Occasional collaborator of CMAP, Ecole Polytechnique.
Recent services
IEEE Tr. PAMI Associate editor (in break)
JMLR action editor .
Member of INRA Evaluation committee since october 2023.
Senior Area Chair for ICML 2023, 22, Neurips 2023 2022, 21
Member of Advising board of IVADO (Montreal, Canada)
Co-program chair at NeurIPS 2019
Senior Area chair at Neurips 2018
Member of program committee of Data Summer School, Palaiseau, May 28 – June 1, 2018
News
November 2024:
. keynote at the Trilateral AI Conference 2024, 12-13 Nov in Tokyo
. Invited talk at a meeting organized by GDR IASIS in Paris about scaling up kernel methods: Nov 22, 2024.
. Invited talk about Graph prediction at IP Paris, Nov 29
September 2024: our paper Any2Graph: Deep End-To-End Supervised Graph Prediction With An Optimal Transport Loss with P. Krzakala, Junjie Yang, R. Flamary, C. Laclau and M. Labeau has just been accepted at NeurIPS 2024 !
June 2024: our paper Deep Sketched Output Kernel Regression for Structured Prediction with Tamim El Ahmad, Junjie Yang and Pierre Laforgue has just been accepted at ECML/PKDD (3) 2024
January 2024: our paper “Sketch in, Sketch out: Accelerating both Learning and Inference for Structured Prediction with Kernels” cosigned with El Ahmad, T., Brogat-Motte, L., Laforgue, P. has been accepted at AISTATS2024 !
December 2023: our paper “Tackling Interpretability in Audio Classification Networks with Non-negative Matrix Factorization” co-signed with Jayneel Parekh, Sanjeel Parekh, Pavlo MOzharovskyi and Gaël Richard has just been accepted for publication by IEEE Tr. Audio, Signal and Language Processing.
October 2023: our paper “Fast Kernel Methods for Generic Lipschitz Losses via p-Sparsified Sketches” co-signed with Tamim El Ahmad and Pierre Laforgue has been accepted to TMLR journal
18 September 2023: keynote at SCEFA workshop, joint to ECML/PKDD 2023, Turin (Polito).
6-8 September 2023: Kick-off Meeting of ELIAS: European Lighthouse in AI & Sustainability
15 June 2023: talk @ Hi! Paris Workshop on frugal AI
19 June 2023: talk +organization of Workshop Frugality in Machine Learning, co-organized with AIRBUS Defense & Space in the context of the Télécom Paris research and teaching DSAIDIS Chair
27-28 June 2023: talk @ the Distance-Based Methods In Machine Learning @ UCL, UK.
APRIL 2023: Project ELIAS “European Lighthouse on Artificial Intelligence & Sustainability” is accepted (coordinator: Nicu Sebe, Trento University).
MARCH 2023: our paper Wind power predictions from nowcasts to 4-hour forecasts: a learning approach with variable selection, by Dimitri Bouche, Rémi Flamary, Florence d’Alché-Buc, Riwal Plougonven, Marianne Clausel, Jordi Badosa and Philippe Drobinski published in Renewable energy.
Previously:
SEPT 2022: our paper about explainability of audio recognition systems: Listen to Interpret: Post-hoc Interpretability for Audio Networks with NMFJ Parekh, S Parekh, P Mozharovskyi, F d’Alché-Buc, G RichardarXiv preprint arXiv:2202.11479, has just been accepted at NeurIPS 2022.
- MAY 2022: We have 2 papers @ ICML 2022: “Learning to Predict Graphs with Fused Gromov-Wasserstein Barycenters”, Luc Brogat-Motte, Rémi Flamary, Céline Brouard, Juho Rousu + “Functional Output Regression with Infimal Convolution: Exploring the Huber and epsilon-insensitive Losses”, With Alex Lambert, Dimitri Bouche and Zoltan Szabò.
Colloquium @ ENS, April 8, 2022.
Keynote @ LIKE22: January 2022
Keynote at Mathias 2021
Keynote paper at CAP 2021: Structured prediction: a kernel view
Our papers: Nonlinear Functional Output Regression: A Dictionary Approach and When OT meets MoM: Robust estimation of Wasserstein Distance are accepted at AISTATS 2021.
Tutorial: Kernels: shallow and deep lerning, Data Science Summer School, Jan 6-10 2021
Keynote address: Kernel-based Structured Output Prediction for Metabolite Identification, Machine Learning for Science and Engineering Conference Dec 14-15, 2020, Columbia University.
our paper Duality in RKHSs with Infinite Dimensional Outputs: Application to Robust Losses accepted @ ICML 2020our paper Audio-Based Detection of Explicit Content in Music accepted @ ICASSP 2020
- Talk @ Parietal, INRIA, July 21, 2020
- AI & Genomics: talk at Colloque Génomique Numérique, nov 21, 2019.
- our ACML paper about Functional Isolation Forest to be presented by Guillaume Staerman in Nagoya, November 19-21, 2019
- our EMLNP paper to be presented by Pierre Colombo in Hong-Kong, November 3-7, 2019
– Talk about AI, Sept 16, Airbus, Paris, 2019.
– Talk about AI, March 27, Académie de Versailles, Journée ISN, INRIA Rocquencourt, 2019
– Jan 2019: Our papers “Infinite task Learning with RKHSs” with Romain Brault and Alex Lambert (first authors) and Zoltan Szabo and Maxime Sangnier, and “Autoencoding any data with kernel autoencoders” with Pierre Laforgue (first author) and Stephan Clémençon just accepted at AISTATS 2019
– Talk at Horizon Maths about AI and ML, Nov 23, 2018.
– Talk at Journées France-IA, Oct 17, 2018.
– Member of Programme Committee of Workshop on Machine Learning and Explainability, Orléans, Oct 8.
– Meetup Women in Machine Learning, Sept 26, 2018.
– Talk at Observatoire AAF du digital, Sept 19, 2018: Artificial Intelligence and Machine Learning, successes and challenges.
– Sept 18, 2018: Our paper “Temporal clustering analysis of endothelial cell gene expression following exposure to a conventional radiotherapy dose fraction using Gaussian process clustering” with Markus Heinonen and Olivier Guipaud and colleagues from IRSN has just been accepted to Plos one.
– Organizing Int. Workshop on Machine Learning and AI, Chair Machine Learning for Big Data, Télécom Paris, Paris, 17-18 Sept 2018.
– Member of program committee of 3rd Junior Conference on Data Science and Engineering, Orsay, 13-14 Sept 2018.
– Talk at Workshop on Gaussian Processes, Sheffield University, Sept 6, 2018.
– Our paper A structured Prediction Approach for Label Ranking accepted at NIPS 2018.
– Déjeuner scientifique aux Journées Statistiques 2018 (31 Mai 2018)
– Our paper on Structured Output Learning with Abstention, accepted at ICML 2018.