Emmanuel J. Candès
Also known as Emmanuel Candes, Emmanuel Candès
Emmanuel Candès is a Stanford mathematician and statistician who develops methods to put statistical guarantees on AI predictions. He completed his statistics PhD at Stanford in 1998 and returned in 2009 after a faculty post at Caltech. His compressed sensing work developed ways to recover images from sparse measurements, and he later coauthored methods for scientific inference with machine learning and for checking language-model outputs. His honours include the 2006 Waterman Award, a 2017 MacArthur Fellowship, the 2020 Princess of Asturias Award for Technical & Scientific Research, and a shared 2026 Shaw Prize in Mathematical Sciences. As of October 2026, he holds Stanford's Barnum-Simons Chair in Mathematics and Statistics, serves as faculty director of Stanford Data Science, and is a senior fellow at the Stanford Institute for Human-Centered Artificial Intelligence.
- Now at
- Stanford University
- Works
- 332
- Citations
- 124,672
- h-index
- 104
- Rankings
- #50 by citations
Publication numbers from OpenAlex, as of 27 October 2026.
Timeline
-
1993–1998
Engineering and a statistics doctorate
He earns an engineering degree at École Polytechnique in 1993 and an MSc at Université de Paris VI in 1994. At Stanford, he completes a statistics PhD in 1998 with Donoho as his advisor and “Ridgelets: Theory and Applications” as his dissertation.
-
1998–2009
Stanford and Caltech faculty posts
He serves as an assistant professor of statistics at Stanford from 1998 to 2000, then joins Caltech in 2000. In 2006, he takes the Ronald and Maxine Linde Chair in Applied and Computational Mathematics. He returns to Stanford as a professor of mathematics and statistics in 2009.
-
2006
Signal recovery and the Waterman Award
In 2006, he receives the Alan T. Waterman Award as a professor of applied and computational mathematics in Caltech's Division of Engineering and Applied Science. His sampling methods recover signals and images from a small amount of data.
-
2014–2018
Academy elections and department leadership
In 2014, he is elected to the National Academy of Sciences and the American Academy of Arts and Sciences. He chairs Stanford's Department of Statistics from 2015 to 2018 and receives a MacArthur Fellowship in 2017. His work includes compressed sensing, which reconstructs images from sparse measurements, and matrix completion, which estimates missing entries in data tables.
-
2020
A shared Princess of Asturias Award
He shares the 2020 Princess of Asturias Award for Technical & Scientific Research with Yves Meyer, Ingrid Daubechies and Terence Tao. The award recognises their contributions to mathematical data and signal processing, including wavelets, compressed sensing and matrix completion.
-
2023
Scientific advice for Fairgen
In 2023, he serves as a scientific advisor to Fairgen, which uses generative AI to augment real data. He meets its engineers weekly to work on technical problems and strategy, with an interest in rigorous validation and avoiding bias.
-
September 2023
Cross-prediction-powered inference
In September 2023, he and Tijana Zrnic release “Cross-Prediction-Powered Inference.” The method combines a small labelled dataset with a large unlabelled dataset, uses machine learning to predict missing labels, and corrects prediction errors to support valid statistical inference.
-
2024
Validity guarantees for language models
In 2024, he coauthors “Large language model validity via enhanced conformal prediction methods” with John J. Cherian and Isaac Gibbs at NeurIPS. They develop methods that give statistical guarantees for language-model outputs and test them on biography and medical question-answering datasets.
-
2026
A shared Shaw Prize
In 2026, he shares the Shaw Prize in Mathematical Sciences with Camillo De Lellis. Candès is recognised for using mathematical analysis to address applied problems in information theory, signal processing and statistics.
-
As of October 2026
Mathematics and statistics at Stanford
As of October 2026, he holds the Barnum-Simons Chair in Mathematics and Statistics at Stanford, where he is a professor of mathematics and statistics. He is faculty director of Stanford Data Science and a senior fellow at the Stanford Institute for Human-Centered Artificial Intelligence. His other appointments include professor of electrical engineering by courtesy and membership of the Institute for Computational and Mathematical Engineering.
Career
- Since 2012Stanford UniversityBarnum-Simons Chair in Mathematics and Statistics
- Since 2009Stanford UniversityProfessor of mathematics and statistics
- NowStanford UniversityFaculty director of Stanford Data Science
- NowStanford UniversitySenior fellow
- NowStanford UniversityProfessor of electrical engineering by courtesy
- NowStanford UniversityMember
- 2025–2026BlackRock
- 2015–2018Stanford UniversityChair
- 2008–2011University of California, San Diego
- 2006–2011California Institute of TechnologyRonald and Maxine Linde Chair in Applied and Computational Mathematics
- 2000–2009California Institute of TechnologyLecturer, then associate professor and professor
- 1998–2000Stanford UniversityAssistant professor of statistics
- California Institute of TechnologyProfessor of applied and computational mathematics
Education
- 1998Stanford UniversityPhD in statistics
- 1994Sorbonne UniversityMSc
- 1993École PolytechniqueEngineering degree
Sources and credits
Photo: We have no published photo of Emmanuel J. Candès yet. Send us a free-licensed photo.
- Emmanuel Candès - MacArthur Foundation
- Emmanuel J. Candès | Department of Statistics
- Emmanuel Candès (X90), co-winner of the 2026 Shaw Prize in Mathematical Sciences - École polytechnique
- Candes Receives Waterman Award - www.caltech.edu
- Emmanuel Candes' Profile | Stanford Profiles
- Yves Meyer, Ingrid Daubechies, Terence Tao and Emmanuel Candès - Fundación Princesa de Asturias
- A Conversation with Fairgen's Scientific Advisor, Emmanuel Candès | Fairgen
- [2309.16598] Cross-Prediction-Powered Inference
- Large language model validity via enhanced conformal prediction methods

