Section 17. Statistics and Data Analysis

All areas of statistics, including inference, parametric and nonparametric statistics, together with all branches of mathematics for data science, where data science includes machine learning, signal and image processing, data generation, data representation, and their applications.
Francis Bach

INRIA Paris, France

Francis Bach is a researcher at Inria in Paris, France, in the Computer Science Department of École Normale Supérieure within PSL Research University. He graduated from Ecole Polytechnique in 1997 and completed his Ph.D. in Computer Science at U.C. Berkeley in 2005. He was elected to the French Academy of Sciences in 2020.

His research interests include machine learning, optimization and statistics, with a particular focus on the design and analysis of efficient algorithms for large-scale problems in data science.

Bin Dong

Peking University, China

Bin Dong is a faculty member of the Beijing International Center for Mathematical Research at Peking University. He received his B.S. from Peking University in 2003, M.Sc from the National University of Singapore in 2005, and Ph.D. from the University of California Los Angeles in 2009. He received the Qiu Shi Outstanding Young Scholar Award in 2014. Bin Dong’s research interest is in the mathematical foundations of image and data analysis and its applications. This includes mathematical analysis, modeling and computations in image processing, medical imaging, and deep learning. In particular, he and his collaborators designed efficient algorithms for medical image reconstruction and segmentation, provided mathematical analysis connecting the PDE-based and wavelet-based approaches to image processing, which also leads to a mathematical understanding of deep neural networks.
Stefanie Jegelka

MIT, USA

Stefanie Jegelka is an Associate Professor in the Department of Electrical Engineering and Computer Science at MIT and a member of the Computer Science and Artificial Intelligence Laboratory. She received her Dr.sc. from ETH Zurich and the Max Planck Institute for Intelligent Systems in Tuebingen, Germany, and also spent time as a postdoctoral researcher at UC Berkeley. She has received a Sloan research fellowship, an NSF CAREER award, a DARPA Young Faculty Award and a Best Paper Award at the International Conference on Machine Learning (ICML) in 2013. In 2022, she is serving as a program chair for ICML, one of the largest machine learning conferences.

Her research interests span modeling and algorithm design for machine learning. In particular, she is known for her work on machine learning with discrete and combinatorial structure. This includes leveraging the interplay between discrete and continuous optimization, such as modeling with and optimization of submodular functions in machine learning. Recently, she has been studying theoretical foundations of neural networks for graph data. Her other research interests include robustness in machine learning, learning with limited labels, negative dependence and its connections, and understanding (learned) data representations.

Yann LeCun

Facebook AI Research and Courant nstitute of Mathematical Sciences, New York University, USA

Lecture on some of the mathematical questions raised by deep learning

Jointly in sections 14, 18

Yann LeCun is VP & Chief AI Scientist at Facebook and Silver Professor at NYU affiliated with the Courant Institute of Mathematical Sciences & the Center for Data Science.

He was the founding Director of Facebook AI Research and of the NYU Center for Data Science. He received an Engineering Diploma from ESIEE (Paris) and a PhD from Sorbonne Université.

After a postdoc in Toronto he joined AT&T Bell Labs in 1988, and AT&T Labs in 1996 as Head of Image Processing Research.

He joined NYU as a professor in 2003 and Facebook in 2013. His interests include AI machine learning, computer perception, robotics and computational neuroscience.

He is the recipient of the 2018 ACM Turing Award (with Geoffrey Hinton and Yoshua Bengio) for «conceptual and engineering breakthroughs that have made deep neural networks a critical component of computing», a member of the National Academy of Engineering and a Chevalier de la Légion d’Honneur.

Oleg V. Lepski

Aix-Marseille Universite, France

Oleg V. Lepski is a Professor of Mathematics at the Mathematical Departement of Aix-Marseille University. He is also a member of the Marseille Mathematical Institute.

His main research interests are related to the minimax theory in non-parametric statistics and they cover almost all areas of current research: non-parametric estimation, adaptive theory, theory of hypothesis testing and statistical inverse problems. Oleg Lepski is the author of the method in minimax adaptive theory which bears his name. He is also the author of several probabilistic papers on uniform estimates of functionals of empirical and Gaussian processes.

Oleg Lepski was awarded Medallion Lecture by the Institute of Mathematical Statistics (2005) and Le Cam’s Lecture by the French Statistical Society (2019).

Gabor Lugosi

Pampeu Fabra University, Spain

Gabor Lugosi is an ICREA research professor at the Department of Economics and Business, Pompeu Fabra University, Barcelona. He received his Ph.D. from the Hungarian Academy of Sciences in 1991. His research has mostly focused on the mathematical aspects of machine learning and related topics in probability and mathematical statistics, including combinatorial statistics, the analysis of random structures, and information theory. He is a co-author of several monographs on pattern recognition, density estimation, online learning, and concentration inequalities.
Richard Nickl

Cambridge University, UK

Richard Nickl is a Professor of Mathematical Statistics at the University of Cambridge, UK.

His research is concerned with the mathematical theory of high- and infinite-dimensional statistical models and random structures that are commonly encountered in modern data science.

His specific recent interests are rigorous statistical and computational guarantees for Bayesian inference algorithms used in non-linear inverse problems arising with partial differential equations.

He is co-author of the award-winning 2016 monograph `Mathematical foundations of infinite-dimensional statistical models’ published by Cambridge University Press. He has received various honours for his work, including the 2017 Ethel Newbold Prize and a 2015 ERC consolidator grant. In 2020 he was an invited speaker at the European Congress of Mathematicians.

Bernhard Schölkopf

MPI for Intelligent Systems, Germany

Bernhard Schölkopf is a Director at the Max Planck Institute for Intelligent Systems (Tübingen, Germany).

His scientific interests are in machine learning and causal inference, including applications to fields ranging from biomedical problems to computational photography and astronomy.

He is a member of the German Academy of Sciences (Leopoldina), has (co-)received the Academy Prize of the Berlin-Brandenburg Academy of Sciences and Humanities, the Royal Society Milner Award, the Leibniz Award, the Koerber European Science Prize, and the BBVA Foundation Frontiers of Knowledge Award. He is an ACM Fellow, a CIFAR Fellow and a Professor at ETH Zürich.

David Silver

Google DeepMind, UK

Lecture on recent breakthroughs in reinforcement learning

Also in section 14

David Silver is a distinguished research scientist at DeepMind and a professor at University College London.

David’s work focuses on artificially intelligent agents based on reinforcement learning. David co-led the project that combined deep learning and reinforcement learning to play Atari games directly from pixels (Nature 2015).

He also led the AlphaGo project, culminating in the first program to defeat a top professional player in the full-size game of Go (Nature 2016), and the AlphaZero project, which learned by itself to defeat the world’s strongest chess, shogi and Go programs (Nature 2017, Science 2018).

Most recently he co-led the AlphaStar project, which led to the world’s first grandmaster level StarCraft player (Nature 2019).

His work has been recognised by the Marvin Minsky award, Mensa Foundation Prize, and Royal Academy of Engineering Silver Medal.

Cun-Hui-Zhang

Rutgers University, France

Cun-Hui Zhang is a Distinguished Professor of Statistics at Rutgers University, USA. His research interests include high-dimensional data, empirical Bayes, nonparametric and semiparametric inference, among other topics. In particular, he is known for his work on regularized estimation with the minimax concave penalty, de-biased inference with high-dimensional data, deconvolution and general maximum likelihood empirical Bayes, and doubly censored data. He is a Fellow of the Institute of Mathematical Statistics and a Fellow of the American Statistical Association.
Tue Oct 05 2021 16:09:31 GMT+0300 (Moscow Standard Time)