Margaux Zaffran
Since November 2024, Margaux is a postdoctoral researcher at UC Berkeley, working with Ryan J. Tibshirani and Aaditya Ramdas. She obtained her PhD in June 2024 from École polytechnique (Paris). She was supervised by Aymeric Dieuleveut (École polytechnique) and Julie Josse (Inria). During her PhD, she has worked closely with Electricité de France, the leading producer and supplier of electricity in France and Europe, in collaboration with Olivier Féron and Yannig Goude. Her research interests revolves around predictive uncertainty quantification (UQ) in general. She likes to work on problems motivated by, or connected with, societal applications, such as energy, climate, or healthcare. Her most important achievements focus on conformal prediction methods. During her PhD, she has extended this class of method to the online setting, which breaks the key i.i.d. assumption, motivated by a collaboration with Electricté de France. She has also studied the impact of missing covariates on predictive UQ and transferred the knowledge to the public-private healthcare consortium TraumaTrix.



Mathilde Mougeot
Mathilde Mougeot is Professeur of Data Science at Ecole Nationale Supérieure d'Informatique pour l'Industrie et l'Entreprise (ENSIIE) and adjunct Professor at ENS Paris Saclay where she holds the Industrial Research Chair "Industrial Data Analytics & Machine Learning". Her research focuses mainly on scientific issues related to predictive models in various contexts, such as those of high dimensionality, model aggregation, domain adaptation, data frugality by model transfer or by hybrid models. Since the beginning of her career, Mathilde Mougeot has been interested in machine learning for artificial intelligence applications. Her research activity is motivated by questions related to concrete applications stemming from collaborative projects with the socio-economic world. She offers a strong experience in leading projects at the interface of academics and industry.



Antoine de Mathelin
Dr. Antoine de Mathelin is a researcher in machine learning, specializing in domain adaptation, transfer learning, uncertainty quantification, and active learning. He completed a PhD focused on developing reliable machine learning models under constraints inherent to engineering design, such as domain shift and limited labeled data. He worked on different aspects related to importance weighting, active learning, and Bayesian neural networks, usually in an high-dimensional context with applications in engineering design. He is also the maintainer of the Adapt library, a useful tool for implementing transfer learning and domain adaptation techniques in Python.



Mounir Atiq
Dr. Mounir Atiq is a post-doctoral researcher working with CEA on graph signal processing for atmospheric surveillance. During his PhD at Centre Borelli, ENS Paris-Saclay, he worked on transfer learning and budgeted learning on Random forest models applied to a fall detection floor system designed by Tarkett company. This research resulted in tools to adapt random forests trained on simulated data with unlimited resources to real time prediction on real-world data using a resource-limited embedded system. He contributed to the transfer learning Python library ADAPT on decision tree based algorithms.