The Schedule

Day 1

August 12th

Eastern Daylight Time in the US

08:20 - 08:30

Kickoff presentation (video)

Gregery Buzzard, Professor of Mathematics at Purdue University

08:30 - 09:20

Thoughts on Deep Learning (Abstract) (video)

Ronald DeVore, the Walter E. Koss Professor and a Distinguished Professor of Mathematics at Texas A&M University

09:20 - 10:10

Deep Approximation via Deep Learning (Abstract) (video)

Zuowei Shen, Tan Chin Tuan Centennial Professor of Mathematics at the National University of Singapore


Break 10:10 - 10:20

10:20 - 11:10

Training and analysis of numerical PDE by neural networks (Abstract) (video)

Jinchao Xu, the Verne M. Willaman Professor Mathematics at the Pennsylvania State University

11:10 - 12:00

A Priori Error Analysis for Solving High-Dimensional PDEs based on Neural Networks (See Abstract)

Jianfeng Lu, Professor of Mathematics, Chemistry, and Physics at Duke University

Lunch 12:00 - 13:30

13:30 - 14:20

Interpreting Deep Neural Networks (Abstract) (video)

Bin Yu, Chancellor's Professor of Statistics, Electrical Engineering, and Computer Sciences at the University of California, Berkeley

14:20 - 15:10

Minimum Complexity Interpolation in Random Features Models (See Abstract)

Andrea Montanari, Professor of Electrical Engineering, Statistics, and Mathematics at Stanford University

Break 15:10 - 15:20

15:20 - 16:10

Computational Lower Bounds for Tensor PCA (See Abstract)

Daniel Hsu, Associate professor of Computer Science at Columbia University

16:10 - 17:00

On the Foundations of Deep Learning: SGD, Overparametrization, and Generalization (See Abstract)

Jason Lee, Assistant professor of Electrical and Computer Engineering and Computer Science at Princeton University

Dinner 17:00 - 19:00

Contributed Talks 19:00 - 21:30

See Abstract via [link]

19:00-19:20: Towards Neural Network Approaches for High Dimensional Problems, Mo Zhou, Duke University, (video)

19:20-19:40: Optimization with Learning-Informed Partial Differential Equation Constraints, Guozhi Dong, Humboldt University of Berlin (video)

19:40-20:00: Solving PDEs on Unknown Manifolds with Machine Learning, Senwei Liang, Purdue University (video)

20:00-20:20: Deep Ritz Method for the Spectral Fractional Laplacian Equation using the Caffarelli-Silvestre Extension, Yiqi Gu, The University of Hong Kong

10 Minute Break

20:30-20:50: Physics-Informed Neural Networks for Shear-Induced Particle Migration, Daihui Lu, Purdue University (video)

20:50-21:10: A Reduced Order Schwarz Method for Nonlinear Multiscale Elliptic Equations based on Two-Layer Neural Networks, Shi Chen, University of Wisconsin-Madison (video)

21:10-21:30: A Deep Learning Based Discontinuous Galerkin Method for Hyperbolic Equations with Discontinuous Solutions and Random Uncertainties, Liyao Lyu, Michigan State University (video)

Day 2

August 13th

Eastern Daylight Time in the US

08:30 - 09:20

Approximating Functions, Functionals, and Operators Using Deep Neural Networks for Diverse Applications (Abstract) (video)

George Karniadakis, the Charles Pitts Robinson and John Palmer Barstow Professor of Applied Mathematics and Engineering, Brown University

09:20 - 10:10

Scalable Algorithms for Bayesian Deep Learning via Stochastic Gradient Monte Carlo and Beyond (Abstract) (video)

Guang Lin, Professor of Department of Applied Mathematics & School of Mech. Eng., Director of Data Science Consulting Service, Purdue University

Break 10:10 - 10:20

10:20 - 11:10

Equivariance and Fourier Space Neural Networks (See Abstract)

Risi Kondor, Associate Professor of Computer Science, Statistics, and Computational and Applied Mathematics, The University of Chicago

11:10 - 12:00

A Layer-Parallel Approach for Training Deep Neural Networks (Abstract) (video)

Eric C. Cyr, Principal Member at the Computer Science Research Institute at Sandia National Laboratories

Lunch 12:00 - 13:30

13:30 - 14:20

Enabling Zero-Shot Generalization in AI4Science (Abstract) (video)

Anima Anandkumar, Bren Professor of Computer & Mathematical Sciences, California Institute of Technology

14:20 - 15:10

The Sobolev Regularization Effect of Stochastic Gradient Descent (Abstract) (video)

Lexing Ying, Professor of Mathematics, Stanford University

Break 15:10 - 15:20

15:20 - 16:10

Deep Networks and the Multiple Manifold Problem (See Abstract)

John Wright, Associate Professor of Electrical Engineering at Columbia University

16:10 - 17:00

Geometry of Linear Convolutional Networks (Abstract) (video)

Guido Montúfar, Assistant Professor of Mathematics and Statistics, University of California, Los Angeles

Dinner 17:00 - 19:00

Contributed Talks 19:00 - 21:30

See Abstract via [link]

19:00-19:20: Galerkin Transformer, Shuhao Cao, Washington University in St. Louis (video)

19:20-19:40: Approximation Properties of Deep ReLU CNNs, Juncai He, The University of Texas at Austin (video)

19:40-20:00: Deep Networks and the Multiple Manifold Problem, Sam Buchanan, Columbia University (video)

20:00-20:20: Overparameterization of Deep ResNet: Zero Loss and Mean-Field Analysis, Zhiyan Ding, University of Wisconsin-Madison (video)

10 Minute Break

20:30-20:50: Asymptotic Optimality and Minimal Complexity of Classification by Random Projection, Evzenie Coupkova, Purdue University (video)

20:50-21:10: Multi-Objective Reinforcement Learning with Non-Linear Scalarization, Mridul Agarwal, Purdue University (video)