Ahmed El Alaoui
Ahmed El Alaoui is a statistics and machine-learning researcher whose work connects high-dimensional inference, algorithms, and statistical physics. He studied applied mathematics at École Polytechnique, completed an M.Sc. in mathematics, vision and learning, and earned a Ph.D. at UC Berkeley in 2018. His papers developed randomized kernel methods and studied spike detection. Later, he connected free-energy methods with computational-hardness theory and proposed a statistical-physics-inspired hierarchy for tensor PCA. He joined Cornell in 2021 and is an assistant professor of statistics and data science there.
- Now at
- Cornell University
Timeline
-
2009–2013
Applied math and learning degrees
He studies applied mathematics at École Polytechnique and earns an engineering diploma in 2012. From 2012 to 2013, he completes an M.Sc. in Mathématiques, Vision et Apprentissage at École Normale Supérieure and École des Ponts ParisTech.
-
November 2014
Randomized kernel ridge regression
In November 2014, he and Michael W. Mahoney post Fast randomized kernel ridge regression with statistical guarantees. Their approach builds a small sketch of the input for faster kernel learning.
-
2018
A Berkeley doctorate
In 2018, he earns a Ph.D. in electrical engineering and computer sciences from UC Berkeley, advised by Michael I. Jordan.
-
2018
Spiked matrix detection limits
With Michael I. Jordan, he studies detection of a single spike in high-dimensional rectangular data in a 2018 COLT paper. The paper analyzes the likelihood ratio and reports asymptotically Gaussian fluctuations in a region below the BBP threshold.
-
November 2020
A Richard M. Karp fellowship
In November 2020, he is a Richard M. Karp Research Fellow at UC Berkeley's Simons Institute for the Theory of Computing.
-
By 2021
A Stanford postdoc
By 2021, he works as a postdoctoral researcher at Stanford, hosted by Andrea Montanari.
-
2022
A NeurIPS paper on hardness
With Afonso S. Bandeira, Samuel Hopkins, Tselil Schramm, Alexander S. Wein, and Ilias Zadik, he publishes The Franz-Parisi criterion and computational trade-offs in high-dimensional statistics at NeurIPS in 2022. The paper connects free-energy methods to low-degree hardness for high-dimensional inference.
-
2025
Kikuchi algorithms for tensor PCA
With Alexander S. Wein and Cristopher Moore, he proposes a hierarchy of increasingly powerful algorithms for tensor PCA, inspired by statistical physics. The paper appears in the Journal of the ACM in 2025.
-
Since 2021
Assistant professor at Cornell
He joins Cornell in 2021 as an assistant professor of Statistics and Data Science and focuses on high-dimensional statistics, machine learning, algorithms, and statistical physics.
Career
- Since 2021Cornell UniversityAssistant Professor of Statistics and Data Science
- University of California, BerkeleyRichard M. Karp Research Fellow
- Stanford UniversityPostdoctoral researcher
Education
- 2018University of California, BerkeleyPh.D. in electrical engineering and computer sciences
- 2013École Normale SupérieureM.Sc. in Mathématiques, Vision et Apprentissage
- 2013École des Ponts ParisTechM.Sc. in Mathématiques, Vision et Apprentissage
- 2012École PolytechniqueEngineering diploma
Sources and credits
Photo: We have no published photo of Ahmed El Alaoui yet. Send us a free-licensed photo.
- Ahmed El Alaoui
- Fast Randomized Kernel Ridge Regression with Statistical Guarantees
- Ahmed El Alaoui | Department of Computer Science | Cornell Bowers
- Detection limits in the high-dimensional spiked rectangular model
- From the Inside: Probability, Geometry, and Computation in High Dimensions
- Ahmed El Alaoui | Cornell Bowers
- The Franz-Parisi Criterion and Computational Trade-offs in High Dimensional Statistics
- The Kikuchi hierarchy and tensor PCA - MaRDI portal
- The Kikuchi Hierarchy and Tensor PCA
- Ahmed El Alaoui | Department of Statistics & Data Science | Cornell Bowers

