Find Academic Supervisors

Francis Bach is a Professor and researcher at INRIA, leading the SIERRA project-team since 2011, which is part of the Computer Science Department at Ecole Normale Supérieure (ENS) within PSL Research University. His work bridges CNRS, ENS, and INRIA as a joint research effort. Elected to the French Academy of Sciences in 2020, he currently runs the ERC project SEQUOIA following his previous ERC project SIERRA (2009-2014). His research spans statistical machine learning with focus on optimization, sparse methods, kernel-based learning, neural networks, graphical models, and signal processing. Bach completed his Ph.D. in Computer Science at U.C. Berkeley under Professor Michael Jordan, followed by work at Ecole des Mines de Paris and the WILLOW project-team at INRIA/ENS/CNRS (2007-2010). His recent book "Learning Theory from First Principles" was published by MIT Press in December 2024. Bach's publication record shows consistent high-impact contributions across machine learning theory and applications, with recent work focusing on conformal prediction, diffusion models, optimization theory, and learning theory foundations. His research demonstrates strong connections between theoretical guarantees and practical algorithms, with applications spanning generative modeling, robust optimization, and statistical inference. Elected to French Academy of Sciences (2020) ERC project SIERRA (2009-2014) ERC project SEQUOIA (current) Author of "Learning Theory from First Principles" (MIT Press, 2024) Bach actively mentors numerous PhD students and postdocs, with many alumni now holding faculty positions at institutions like EPFL, Ecole Polytechnique, University of Washington, and University of Montreal. His teaching includes advanced courses on learning theory at ENS's Master's programs. He regularly presents tutorials at major conferences including COLT, NeurIPS, and ICML, demonstrating his leadership in the theoretical machine learning community.

• statistical machine learning
• optimization
• sparse methods
• +5 more

Florence d'Alché-Buc is a Professor at Télécom Paris (Institut Polytechnique de Paris), holding an Isaac Newton Institute Simons Chair (2025) and leading the Data Science and Artificial Intelligence for Digitalized Industry & Services (DSAI) Chair. She heads the Image, Data, and Signal Department and is part of the Signal, Statistics, and Learning (S2A) team at the LTCI laboratory. Her research focuses on machine learning, bioinformatics, and industrial applications, emphasizing kernel methods, structured prediction, and reliable AI. Education: Previously a professor at Université d’Evry and deputy director of the IBISC lab. Co-director of the Paris-Saclay Data Science Master and creator of specialized AI programs (e.g., Certificate of Specialized Studies in AI). Research highlights include contributions to operator-valued kernel methods, graph prediction, and frugal AI. She actively collaborates with institutions like Inria, École Polytechnique, and industry partners (Airbus, Engie, etc.). Notable roles: Scientific director of Digicosme Labex, Ellis Fellow, and board member of IVADO (Montreal). Her recent work addresses AI explainability, robustness, and sustainability, including projects on interpretable networks and energy-efficient models.

• Machine Learning
• Artificial Intelligence
• Bioinformatics
• +9 more

Alastair F. Donaldson is a Professor and Director of Research in the Department of Computing at Imperial College London, where he leads the FastPL research group. His work bridges formal methods, software testing, and programming languages, with a focus on enhancing the reliability of high-performance and parallel software systems. He has held key roles including Director of Research (since 2023) and previously served as Lecturer (2011–2014), Senior Lecturer (2014–2017), and Reader (2017–2020) before being promoted to Professor in 2020. His research interests include formal verification, compiler testing, GPU programming, concurrency, and fuzzing. He has made significant contributions to the verification of GPU kernels, metamorphic testing of graphics drivers, and the development of tools like GPUVerify and GraphicsFuzz. His work combines theoretical rigor with practical impact, demonstrated by the acquisition of his startup GraphicsFuzz by Google in 2018 and his subsequent roles as Senior Software Engineer and Visiting Researcher at Google. His recent publications reflect a sustained focus on compiler and system reliability, with trends in fuzzing, formal specification, and automated testing of complex systems such as WebGPU, CXL cache coherence, and large language models for code generation. His work increasingly integrates empirical validation with formal techniques to uncover subtle bugs in real-world systems. Scientific awards and recognitions include: 2017 BCS Roger Needham Award EPSRC Early Career Fellowship Fellow of the British Computer Society Best Paper awards at EuroSys 2024, MET 2021, IISWC 2019, IWOCL 2019, and ICST 2016 Best Industry Paper at ICST 2024 ACM SIGSOFT Distinguished Paper at ISSTA 2023 ACM SIGPLAN Most Influential OOPSLA Paper Award (2012 paper), awarded in 2022 Best Student Paper at PPoPP 2014 He has advised numerous PhD students and leads a vibrant research group. He has secured significant research funding and collaborates extensively with industry and academia. His service includes leadership roles such as General Chair of PLDI 2020, PC Chair of ECOOP 2019, and Steering Committee Chair of PLDI (2022–2025). He also serves on the advisory board of PACM-PL and on program committees for top venues including POPL, OOPSLA, PLDI, ICSE, and ISSTA. He leads the FastPL research group, which focuses on the design and implementation of programming tools and techniques for reliable software. The group conducts cutting-edge research in compiler testing, formal methods, and high-performance systems, fostering collaboration across academia and industry.

• Formal Verification
• Software Testing
• Programming Languages
• +12 more

Gang (Gary) Tan is a Professor at the Pennsylvania State University's College of Engineering, specializing in computer security, formal methods, and programming languages. He co-directs the Institute for Networking and Security Research (INSR) and leads the Security of Software (SOS) Group, focusing on compiler, programming language, and formal method techniques to enhance computer security. Education: B.E. in Computer Science from Tsinghua University Ph.D. in Computer Science from Princeton University His research integrates formal verification with practical security applications, particularly emphasizing: Compiler-based security enforcement Side-channel mitigation in speculative execution Fairness analysis in machine learning systems Formal grammar approaches for software reliability Key article trends show: Security-focused formal methods (15% of publications) ML fairness verification (20% of recent work) Compiler-based security solutions (30% of output) Side-channel defense mechanisms (25% of research) Parser design and formal grammar synthesis (10% of contributions) Scientific achievements include: NSF CAREER Award Google Research Awards (2x) PLDI 2024 Best Paper James F. Will Career Development Professorship Outstanding Research Award at Penn State Ruth and Joel Spira Excellence in Teaching Award Dr. Tan actively contributes to academic communities through: DARPA ISAT study group membership Program committee roles (CGO 2024, ECOOP 2018, etc) Leadership in security research initiatives

• Computer Security
• Formal Methods
• Programming Languages
• +2 more

Pierre Marion is a Researcher at INRIA Paris, working within the Sierra research team since September 2025. His work focuses on the theoretical foundations of deep learning and he is beginning to explore applications of AI in mathematics. Marion has established collaborations across multiple institutions including EPFL, Sorbonne Université, and Google DeepMind. His educational background includes: Engineering degree from École polytechnique (2015-2018) with specialization in Applied Mathematics Master's degree from Sorbonne Université (2019-2020) PhD from Sorbonne Université (2020-2023) under the supervision of Gérard Biau and Jean-Philippe Vert Postdoctoral research at EPFL (2024) supervised by Lénaïc Chizat Marion's research interests primarily focus on the theory of deep learning, where he investigates the optimization and statistical properties of various neural network architectures. His work spans from shallow networks to complex generative models, with a particular emphasis on understanding the mathematical foundations that govern deep learning performance. Recently, he has begun exploring applications of AI in mathematical research, aiming to bridge the gap between theoretical machine learning and mathematical discovery. His research often combines rigorous theoretical analysis with practical implications for training deep neural networks. Analysis of Marion's recent publications reveals several key trends in his research. He has made significant contributions to understanding the role of large learning rates in optimization dynamics, demonstrating how they can accelerate convergence in logistic regression and prevent memorization in score-based generative models. His work on attention mechanisms has provided theoretical guarantees for their effectiveness in specific tasks like single-location regression and clustering. Additionally, Marion has extensively studied the connections between residual networks and neural ordinary differential equations , establishing generalization bounds and exploring scaling properties in the large-depth regime. His earlier work included contributions to natural language processing and quasi-Monte Carlo methods, reflecting a broad mathematical foundation that informs his current deep learning research. Marion has received several notable scientific awards: Runner-up PhD Award of AFIA (French Association for Artificial Intelligence) in 2024 Google PhD Fellowship in 2022 Ecole polytechnique Grand Prize of Research Internships in 2018 As an advisor, Marion currently supervises PhD student Yu-Han Wu (since 2024), with whom he has co-authored multiple publications on large learning rates and denoising score matching. Previously, he co-supervised several Master's students including Seorim Park, Yerkin Yesbay, and Nathan Doumèche. Marion has been actively involved in the machine learning community through conference organization (NeurIPS@Paris meetups), session chairing (ICSDS 2022), and extensive reviewing activities. He has served as a reviewer for top journals including JASA and Annals of Statistics, and conferences including NeurIPS and ICLR, where he was recognized as a top reviewer at NeurIPS 2023. Marion is a member of the Sierra research team at INRIA Paris, which focuses on machine learning theory and applications. He has also collaborated with researchers at CREST (Center for Research in Economics and Statistics), as evidenced by his participation in seminars organized by Anna Korba and Karim Lounici. His work often bridges theoretical computer science, statistics, and applied mathematics, reflecting the interdisciplinary nature of modern machine learning research.

• Deep Learning Theory
• Optimization of Neural Networks
• Generative Models
• +2 more

Jaouad Mourtada is an Assistant Professor in the Department of Statistics at ENSAE/CREST, École Nationale de la Statistique et de l'Administration Économique since September 2020. Previously, he was a postdoctoral researcher at the Laboratory for Computational and Statistical Learning at the University of Genoa (2019-2020). He completed his PhD in Statistics at École Polytechnique under the supervision of Stéphane Gaïffas and Erwan Scornet. His educational background includes a Master's degree in Mathematics with specialization in Probability and Random Models (2016), a Master's degree in Fundamental Mathematics (2015), and a Bachelor's degree in Mathematics from Pierre and Marie Curie University and École Normale Supérieure (2013). Dr. Mourtada's research focuses on the intersection of statistics and learning theory, with particular interest in high-dimensional statistics, online learning, and density estimation. His work explores the complexity of prediction and estimation problems through rigorous theoretical analysis. His research spans statistical learning theory, robust statistics, and the theoretical foundations of machine learning algorithms. His publication record since 2017 demonstrates consistent contributions to top-tier venues in statistics and machine learning, with recent work focusing on universal coding, aggregation methods, robust regression, and the theoretical analysis of kernel methods and random forests. His research shows a progression from online learning and expert aggregation to more complex statistical learning problems involving high-dimensional data and model misspecification. He teaches courses including Statistical Learning Theory for Master 2 Data Science students and Probability Theory at ENSAE. His teaching spans theoretical foundations of machine learning and core probability concepts for advanced statistics students.

• Statistics
• Learning Theory
• High-dimensional statistics
• +4 more

Philippe Moireau is a Full Professor in the Department of Applied Mathematics at École Polytechnique, where he is also affiliated with the Center for Applied Mathematics (CMAP). He serves as the head of the Inria Project-Team MΞDISIM (Mathematical and Mechanical Modeling with Data Interaction for Simulation in Medicine) and holds the distinguished position of Ingénieur Général of The Corps des Mines. His primary research focuses on inverse problems and data assimilation for partial differential equation models, with particular emphasis on: Observer-based methods from optimal control perspectives Stabilization approaches for evolution equations Numerical analysis of time-dependent control problems Digital twin applications in cardiovascular medicine Professor Moireau's publication portfolio demonstrates consistent focus on mathematical methods for physical systems, with recurring themes in: Data assimilation techniques for PDE-based models Numerical stabilization and discretization methods Cardiovascular biomechanics and hemodynamics Stochastic modeling of biological systems Epidemiological forecasting and control He leads the ANANKΞ project-team at Inria focused on Analysis And Numerics of physical-Knowledge-based Estimation. His educational contributions include lectures on data assimilation theory at CEMRACS and courses on mathematical modeling in cardiac biomechanics at Institut Polytechnique de Paris.

• Inverse Problems
• Data Assimilation
• Partial Differential Equations
• +4 more

Cristina Butucea is a Professor of Statistics and Machine Learning at ENSAE-CREST, Institut Polytechnique de Paris . She was nominated an IMS Fellow (2019) for her contributions to nonparametric and high-dimensional statistics. She has co-organized major conferences like Fréjus 2018 , Luminy 2019-2020 , and Oberwolfach 2021 , and serves as Associate Editor for ALEA . Fields of interest: Nonparametric statistics, quantum statistics, differential privacy, inverse problems, machine learning. Awards: IMS Fellowship, conference organization leadership. Research trends: Focuses on optimal estimation under privacy constraints, quantum state reconstruction, high-dimensional inference, and adaptive nonparametric methods. Email: Cristina.Butucea@ensea.fr | Cristina.Butucea@ip-paris.fr

• Nonparametric Statistics
• High-Dimensional Statistics
• Inverse Problems
• +3 more

Michalis Vazirgiannis is a Professor at LIX, École Polytechnique in France, where he leads the Data Science and Mining group (DaSciM). He holds a degree in Physics and a PhD in Informatics from Athens University (Greece), and a Master's degree in AI from Heriot Watt University, Edinburgh (UK). His academic career spans multiple prestigious institutions including Fraunhofer and Max Planck MPI in Germany, INRIA/FUTURS in Paris, AUEB in Greece, Telecom-Paristech, ENS in France, Tsinghua and Jiaotong Shanghai in China, and Deusto University in Spain. Professor Vazirgiannis's research focuses on machine and deep learning methods for graph analysis, including community detection, graph clustering, node embeddings, and influence maximization. His work in text mining encompasses Graph of Words, word embeddings with applications to web advertising and marketing, event detection, and summarization. He has active collaborations with industrial partners in analytics and machine learning for large-scale data repositories across various application domains such as recommendations, meeting summarization, influence metrics for scientific and social networks, and predictive maintenance. His recent publications demonstrate a strong emphasis on Graph Neural Networks, multilingual NLP (particularly for French and Arabic), and applications of deep learning to diverse domains including social networks, legal text, and biomedical data. There's a clear trajectory toward developing more efficient, explainable, and specialized models that address real-world challenges in data analysis. ERCIM fellowship Marie Curie EU fellowship Tencent "Rhino-Bird International Academic Expert Award" (2017) Best Paper Award at IJCAI 2018 Best Paper Award at CIKM 2013 Professor Vazirgiannis has supervised 29 completed PhD theses and has attracted significant R&D funding from national and international sources, including research agencies and industrial partners such as Google, Airbus, Huawei, Deezer, BNP, and LVMH. He leads or has led several academic research chairs including DIGITEO (2013-15), ANR/HELAS (2020-25), and AXA (2015-2018). The DaSciM research group, which he leads at École Polytechnique, has extensive experience in real-world R&D projects involving large-scale data mining. The team maintains active collaborations with major industrial partners including AIRBUS, Google, BNP, Tencent, and Tradelab, working on cutting-edge machine learning projects. The group has co-organized major conferences such as ECML PKDD 2011 and ECML/PKDD 2017 and participates in the senior organization of AI and data mining events like AAAI and IJCAI.

• Machine Learning
• Deep Learning
• Graph Analysis
• +11 more

Anna Korba is an Assistant Professor at École Polytechnique, specifically affiliated with ENSAE/CREST in the Statistics Department since September 2020. She is also a co-administrator of the Master Data Science program at École Polytechnique. Her academic journey has positioned her as a leading researcher in machine learning, with particular expertise in kernel methods, optimal transport, and statistical optimization. Dr. Korba received her PhD from Telecom ParisTech in 2018 under the supervision of Prof. Stephan Clémençon. Prior to her current position, she was a postdoctoral researcher at University College London's Gatsby Computational Neuroscience Unit working with Arthur Gretton from December 2018 to August 2020. Her academic foundation includes a Master's degree in Machine Learning and Computer Vision (MVA) from ENS Cachan and ENSAE in 2015. Anna Korba's research primarily focuses on machine learning with emphasis on kernel methods, optimal transport, optimization, particle systems, and preference learning. Her work bridges theoretical statistics with practical machine learning applications, particularly in developing novel sampling and optimization methods. She has made significant contributions to understanding Wasserstein gradient flows, density ratio estimation, and variational inference techniques. Her publication record demonstrates a strong trajectory in top-tier machine learning conferences including ICML, NeurIPS, AISTATS, and ICLR. Her research shows a clear evolution from foundational work on ranking and preference learning during her PhD to more recent contributions in Wasserstein-based optimization, sampling methods, and deep probabilistic modeling. The interdisciplinary nature of her work connects statistics, optimization theory, and practical machine learning applications. Top 10% Oral Presentation at AISTATS 2022 Top 15% Long Oral Presentation at ICML 2021 She actively mentors PhD students and postdoctoral researchers, currently advising seven PhD candidates and having successfully guided several alumni to prestigious positions. Dr. Korba also contributes to the academic community through her role in administering the Master Data Science program and collaborating with researchers across institutions worldwide. As part of the CREST research center, Dr. Korba works within a vibrant team of researchers focused on statistics, machine learning, and their applications to economic and social sciences. Her research group includes current PhD students and postdocs working on various aspects of her research interests, creating a dynamic environment for advancing the field of statistical machine learning.

• Machine Learning
• Nonparametric Statistics
• Kernel Methods
• +5 more

Sylvain Lombardy is a Professor at the University of Bordeaux, affiliated with the Laboratoire Bordelais de Recherche en Informatique (LaBRI) and the Enseirb-Matmeca engineering school within the Institut Polytechnique de Bordeaux. His research focuses on automata theory, formal languages, and theoretical computer science, particularly in weighted automata, formal methods, and algebraic properties of automata. He leads the Formal Methods research team at LaBRI and contributes to projects like the Awali and Vaucanson software platforms for automata manipulation. Education: PhD in Computer Science (2001, ENST Paris), Habilitation à Diriger des Recherches (2005, University of Paris Diderot). His work bridges theoretical foundations with practical tools, emphasizing algorithmic and algebraic aspects of automata. Notable contributions include studies on automata minimization, determinization, and the interplay between rational expressions and automata constructions. Research Interests: Automata Theory, Formal Power Series, Weighted Automata, Tropical Semirings, Algebraic Automata Theory, and Computational Models for Discrete Systems. His recent work explores two-way automata, Hadamard series, and applications in formal verification. Publications highlight contributions to the structure and properties of automata, with a focus on formal methods and algorithmic decidability. Key works address unambiguity, determinism, and the minimization of weighted automata across various semirings. Collaborations include projects on automata-based kernels for machine learning and XML formats for automata descriptions. He has developed influential software tools such as Awali (finite-state machine platform) and Vaucanson (automata manipulation framework), demonstrating practical applications of theoretical research. His work is supported by grants exploring automata in computational linguistics and discrete mathematics.

• Automata Theory
• Formal Languages
• Theoretical Computer Science
• +3 more

Etienne Mémin is a Research Director (Full Professor status) at Inria and leads the Odyssey research group, which is affiliated with multiple institutions including University of Rennes, IRMAR, Ifremer, LOPS, UBO, IMT Atlantique, and Lab-STICC. He serves as a Visiting Professor at the Department of Mathematics, Imperial College London (2020–2026) and is the Principal Investigator of the ERC STUOD grant. His research spans the intersection of geophysical sciences, fluid mechanics, computational sciences, and applied mathematics, focusing on stochastic modeling of fluid flows, data assimilation, and uncertainty quantification. He has developed frameworks for stochastic geophysical flows, coarse-scale simulations, and robust motion estimation techniques. Recent publications highlight his work on stochastic Navier-Stokes equations, ensemble forecasting, and data assimilation for ocean and atmospheric models. He has applied these methods to numerical weather prediction, turbulence analysis, and real-time flow reconstruction using sparse measurements. Scientific Awards: ERC STUOD grant PhD Students: Francesco Tucciarone (ERC STUOD, NEMO code) Benjamin Dufée (Ensemble Kalman filters, ATER position) Berenger Hug (Stochastic Navier-Stokes analysis, teaching) Antoine Moneyron (Stochastic ocean models) Collaborations: Imperial College London (D. Crisan, S. Laizet), Zhejiang University (S. Cai, C. Xu), MétéoFrance (P. Arbogast, O. Pannekoucke), Ifremer (B. Chapron), IRSTEA Lyon (L. Pénard), IRMAR (R. Lewandovsky), University of Buenos Aires (G. Artana), ISSI Beijing (T. Corpetti).

• Stochastic Modeling
• Geophysical Flows
• Fluid Mechanics
• +3 more

Jean-Marc Jezequel is a Professor of Software Engineering at University of Rennes , affiliated with CNRS , Inria , IRISA , and Institut Universitaire de France (IUF) . His research focuses on Model-Driven Engineering , Software Product Lines , Dynamic Adaptation , and Executable Meta-languages . Key Contributions : Pioneering work in aspect-oriented and model-driven approaches for software evolution Foundational research on model transformations (e.g., UMLAUT framework) Advances in testing and validation of distributed systems Research Trends from his recent publications include: Intelligent modeling assistance integrating machine learning Contextual variability modeling for complex systems Runtime model execution for self-adaptive systems Formal methods and constraint resolution for UML validation Collaborations include researchers from Luxembourg, Montreal, Colorado State University, and INRIA.

• Software Engineering
• Model Driven Engineering
• Software Product Lines
• +7 more

Hédi Hadiji is an Associate Professor at the Laboratoire des Signaux et Systèmes (L2S) at CentraleSupélec. His research focuses on the mathematics of online decision-making, particularly adaptivity in bandits and online learning. Prior to this role, he was a postdoctoral researcher at the University of Amsterdam under Tim van Erven. He holds a PhD in Mathematics from Université Paris-Saclay, supervised by Gilles Stoltz and Pascal Massart. Education: PhD in Mathematics, Université Paris-Saclay (2020) Masters in Probability and Statistics, Université Paris-Saclay (2017) Part III Mathematical Tripos, University of Cambridge (2016) École Polytechnique (2012–2016) Research Interests: Hédi’s work addresses theoretical foundations of online learning, including bandit algorithms, reinforcement learning, and optimization. He explores adaptive strategies in dynamic environments, with applications to stochastic and adversarial settings. Key themes include minimax optimal algorithms, regret analysis, and the interplay between exploration and exploitation. Recent Publications: His recent work includes minimax optimal algorithms for linear bandits, tracking solutions in time-varying systems, and diversity-preserving strategies in K-armed bandits. These contributions advance theoretical understanding of adaptive decision-making in complex systems. Teaching: He teaches courses on the theoretical principles of deep learning, reinforcement learning, and mathematical foundations at the graduate level. His lectures emphasize rigorous analysis of generalization, optimization, and the neural tangent kernel. Labs & Affiliations: His primary affiliation is with L2S, where he engages in collaborative research groups such as MODESTY, COMEDY, and SYCOMORE. His work intersects signal processing, control systems, and communication networks.

• Online Decision-Making
• Adaptivity in Bandits
• Online Learning
• +3 more

Maryam Mehri Dehnavi is an Associate Professor in the Department of Computer Science at the University of Toronto and a Principal Research Scientist at NVIDIA. She holds the Canada Research Chair in Parallel and Distributed Computing and leads the ParaMathics research group. Research focuses on high-performance computing , machine learning , sparse matrix optimizations , and compiler design for heterogeneous systems. Her work develops domain-specific languages , scalable numerical libraries , and auto-vectorization techniques for cloud and GPU platforms. Recent publications address LLM compression , sparse code translation , GPU kernel synchronization , and control flow optimization . Scientific recognition: Ontario Early Researcher Award (2021), NSF CRII Grant, NSERC New Frontiers in Research Fund. Current students: Mushegh Shahinyan , Martin Phan , Maryam Haghifam , and others. Former advisees: Kazem Cheshmi (NJIT), Zachary Blanco (MIT Lincoln Lab), Yuanxi Li (Amazon).

• High-Performance Computing
• Cloud Computing
• Machine Learning
• +3 more