Emmanuel Abbé
Professeur ordinaire
EPFL SB MATH MDS1
MA C2 543 (Bâtiment MA)
Station 8
CH-1015 Lausanne
+41 21 693 20 78
+41 21 693 25 50
Local: MA C2 543
EPFL
>
SB
>
MATH
>
MDS1
Web site: Site web: https://mds.epfl.ch/
EPFL CIB
GA 3 34 (Bâtiment GA)
Station 5
CH-1015 Lausanne
+41 21 693 20 78
Local: GA 3 34
EPFL
>
VPA
>
VPA-AVP-CP
>
CIB
>
CIB-GE
Web site: Site web: https://bernoulli.epfl.ch
Mission
Chair of Mathematical Data Science (SB/IC)
The research in the chair of Mathematical Data Science (MDS) focuses on the mathematical principles that underpin the analysis and design of information and data science technologies. As branches of mathematics, this involves probability, statistics, and discrete mathematics, and as specific fields, machine learning and information theory.
Please consult http://deepfoundations.ai/ for postdoc applications.
The research in the chair of Mathematical Data Science (MDS) focuses on the mathematical principles that underpin the analysis and design of information and data science technologies. As branches of mathematics, this involves probability, statistics, and discrete mathematics, and as specific fields, machine learning and information theory.
Please consult http://deepfoundations.ai/ for postdoc applications.
Biographie
Emmanuel Abbe received his Ph.D. degree from the EECS Department at the Massachusetts Institute of Technology (MIT) in 2008, and his M.S. degree from the Department of Mathematics at the Ecole Polytechnique Fédérale de Lausanne in 2003. He was at Princeton University as an assistant professor from 2012-2016 and an associate professor from 2016, jointly in the Program for Applied and Computational Mathematics and the Department of Electrical Engineering, as well an associate faculty in the Department of Mathematics at Princeton University since 2016. He joined EPFL in 2018 as a Full Professor, jointly in the Mathematics Institute and the School of Computer and Communication Sciences, where he holds the Chair of Mathematical Data Science. He is the recipient of the Foundation Latsis International Prize, the Bell Labs Prize, the NSF CAREER Award, the Google Faculty Research Award, the Walter Curtis Johnson Prize, the von Neumann Fellowship from the Institute for Advanced Study, the IEEE Information Theory Society Paper Award, and a co-recipient of the Simons-NSF Mathematics of Deep Learning Collaborative Research Award.
Prof. E. Abbe is also a Global Expert at the Geneva Science and Diplomacy Anticipator (GESDA), a member of the Steering Committee of the Center for Intelligent Systems (CIS), EPFL, and director interim of the College des Sciences, EPFL.
![]("https://people.epfl.ch/private/common/photos/links/120091.jpg")
Publications
Sélection de publications
| E. Abbe, S. Li, A. Sly FOCS'21 |
Proof of the Contiguity Conjecture and Lognormal Limit for the Symmetric Perceptron |
| E. Abbe, P. Kamath, E. Malach, C. Sandon, N. Srebro NeurIPS'21 Spotlight |
On the power of differentiable learning versus PAC and SQ learning |
| E. Abbe, E. Boix-Adsera, M. Brenner, G. Bresler, D. Nagarj NeurIPS'21 |
The staircase property: how hierarchical structure can guide deep learning |
| E. Abbe, E. Cornacchia, Y. Gu, Y. Polyanskiy COLT'21 Best Student Paper Award |
Stochastic block model entropy and broadcasting on trees with survey |
| E. Malach, P. Kamath, E. Abbe, N. Srebro ICML'21 |
Quantifying the Benefit of Using Differentiable Learning over Tangent Kernels |
| A. Asadi, E. Abbe |
Maximum Multiscale Entropy and Neural Network Regularization |
| E. Abbe, J. Fan, K. Wang |
An l_p theory of PCA and spectral clustering |
| E. Abbe, S. Li, A. Sly |
Learning sparse graphons and the generalized Kesten-Stigum Threshold |
| E. Abbe, C. Sandon |
Polytime universality and limitations of deep learning |
| E. Abbe, A. Shpilka, M. Ye Information Theory Trans. |
Reed-Muller codes: theory and algorithms, |
| E. Abbe, C. Sandon NeurIPS 20 |
On the universality of deep learning |
| E. Abbe, M. Ye FOCS 19 |
Reed-Muller codes polarize |
| Min Ye, Emmanuel Abbe ISIT 19 |
Recursive projection-aggregation decoding of Reed-Muller codes |
| Amir R. Asadi, Emmanuel Abbe JMLR |
Chaining Meets Chain Rule: Multilevel Entropic Regularization and Training of Neural Nets |
| Emmanuel Abbe, Enric Boix-Adserà ISIT 19 |
Subadditivity Beyond Trees and the Chi-Squared Mutual Information |
| E. Abbe, E. Boix, P. Ralli, C. Sandon SIAM Journal on Mathematics of Data Science |
Graph powering and spectral robustness |
| E. Abbe, E. Boix Annals of Applied Probability |
An Information-Percolation Bound for Spin Synchronization on General Graphs |
| A. Asadi, E. Abbe, S. Verdu NIPS 18 |
Chaining mutual information and tightening generalization bounds |
| E. Abbe, L. Massoulié, A. Montanari, A. Sly, N. Srivastava Mathematical Statistics and Learning (MSL) 18 |
Group syncrhonization on grids |
| M. Ye, E. Abbe ICML 18 |
Communication-computation efficient gradient coding |
| A. Sankararaman, E. Abbe, F. Baccelli Information and Inference: A Journal of the IMA |
Community Detection on Euclidean Random Graphs |
| E. Abbe, T. Bandory, W. Leeb, J. Pereira, N. Sharon, A. Singer Information Theory Trans. |
Multireference alignment is easier with an aperiodic translation distribution |
| E. Abbe, J. Fan, K. Wang, Y. Zhong Annals of Statistics |
Entrywise eigenvector analysis of random matrices of low expected rank |
| E. Abbe, S. Kulkarni, E. Lee JMLR |
Generalized nonbacktracking bounds on the influence |
| E. Abbe JMLR |
Community detection and stochastic block models: recent development |
| E. Abbe, S. Kulkarni, E. Lee NIPS 17 |
Nonbacktracking bounds on the influence in cascade models |
| E. Abbe, J. Pereira, A. Singer ISIT 16 |
Sample complexity of the Boolean multireference alignment problem |
| E. Abbe, C. Sandon CPAM |
Proof of the achievability conjectures in the general stochastic block model |
| E. Abbe, C. Sandon NIPS 16 oral |
Detection in the stochastic block model with multiple clusters: proof of the achievability conjectures, acyclic BP, and the information-computation gap |
| E. Abbe, C. Sandon NIPS 15 |
Recovering communities in the general stochastic block model without knowing the parameters |
| E. Abbe, C. Sandon FOCS 15 |
Community detection in the general stochastic block model: fundamental limits and efficient recovery algorithms |
| E. Abbe, A. Shpilka, A. Wigderson STOC 15 |
Reed-Muller codes for random erasures and erros |
| Y. Desphande, E. Abbe, A. Montanari Journal Information and Inference: A Journal of the IMA |
Asymptotic mutual information for the balanced binary SBM |
| E. Abbe, Y. Wigderson ISIT 15 |
High-girth matrices and polarization |
| E. Abbe, A. Bandeira, G. Hall Information Theory Trans. |
Exact recovery in the stochastic block model |
| E. Abbe, N. Alon, A. Bandeira ISIT 14 |
Linear Boolean classification, coding and "the critical problem" |
| E. Abbe, A. Bandeira, A. Bracher, A. Singer Transactions on Network Science and Eng. |
Decoding graph labels from censored correlations: phase transition and efficient recovery |
| E. Abbe, A. Montanari Journal Theory of Computing |
Conditional random fields, planted constraint satisfaction and entropy concentration |
| E. Abbe, A. Montanari Random Structures and Algorithms |
On the concentration of the number of solutions of random satisfiability formulas |
| N. Goela, E. Abbe, M. Gastpar Information Theory Trans. |
Polar codes for broadcast channels |
| E. Abbe, A. Khandani, A. W. Lo American Economical Review (AER) |
Privacy-preserving methods in systemic risk |