Scholar
Antoine de Mathelin
Google Scholar ID: h79bffAAAAAJ
ENS Paris-Saclay PhD Student
Citations & Impact
Citations
252
H-index
8
i10-index
5
Publications
17
Co-authors
3
list available
Contact
Publications
Resume (English only)
Academic Achievements
- Personalized One-Shot Collaborative Learning - ICTAI 2023
- From theoretical to practical transfer learning, The ADAPT library - FTL 2022
- Fast and Accurate Importance Weighting for Correcting Sample Bias - ECML-PKDD 2022
- Discrepancy-Based Active Learning for Domain Adaptation - ICLR 2022
- Unsupervised domain adaptation for constraining star formation histories - AI2ASE 2022
- Handling distribution shift in tire design - NeurIPS-DistShift 2021
- A Binded VAE for Inorganic Material Generation - NeurIPS-DGM 2021
- Adversarial weighting for domain adaptation in regression - ICTAI 2021
- Unsupervised multi-source domain adaptation for regression - ECML-PKDD 2020
- Open-source software: ADAPT (Awesome Domain Adaptation Python Toolbox) and SKADA (Domain Adaptation with scikit-learn and PyTorch)
Research Experience
- Works as a Postdoctoral Researcher at the Sloan Kettering Institute for Cancer Research, where he develops deep generative models and active learning algorithms to discover effective combination therapies.
Education
- Earned a Ph.D. from Centre Borelli of the ENS Paris-Saclay, France, under the supervision of Pr. Mathilde Mougeot, Pr. Nicolas Vayatis, and François Deheeger. The thesis was sponsored by the Michelin tire company and focused on developing reliable machine learning models under the intrinsic constraints of engineering design, such as domain shift and costly labeling.
Background
- A Postdoctoral Research Fellow at the Sloan Kettering Institute for Cancer Research, focusing on developing deep generative models and active learning algorithms to discover effective combination therapies. Interested in transfer learning, domain adaptation, active learning, uncertainty quantification, and out-of-distribution detection.
Miscellany
- Committed to translating ML research into real-world applications, particularly in industrial and medical fields.
