| Talk_id | Date | Speaker | Title |
|
29415
|
Friday 9/16
4:00 PM
|
Anna Veselovska, Department of Mathematics, Technical University of Munich, Germany
|
Super-Resolution on the Two-Dimensional Unit Sphere
- Anna Veselovska, Department of Mathematics, Technical University of Munich, Germany
- Super-Resolution on the Two-Dimensional Unit Sphere
- 09/16/2022
- 4:00 PM - 5:00 PM
- C304 Wells Hall
- Mark A Iwen (iwenmark@msu.edu)
In this talk, we discuss the problem of recovering an atomic measure on the unit 2-sphere S^2 given finitely many moments with respect to spherical harmonics. The analysis relies on the formulation of this problem as an optimization problem on the space of bounded Borel measures on S^2 as suggested by Y. de
Castro & F. Gamboa and E. Candes & C. Fernandez-Granda. We construct a dual certificate using a kernel given in an explicit form and make a concrete analysis of the interpolation problem. We support our theoretical results by various numerical examples related to direct solution of the optimization
problem and its discretization.
This is a joint work with Frank Filbir and Kristof Schroder.
|
|
29438
|
Thursday 9/22
2:30 PM
|
Rayan Saab, University of California, San Diego (UCSD)
|
Quantizing neural networks
- Rayan Saab, University of California, San Diego (UCSD)
- Quantizing neural networks
- 09/22/2022
- 2:30 PM - 3:30 PM
- Online (virtual meeting)
(Virtual Meeting Link)
- Mark A Iwen (iwenmark@msu.edu)
Neural networks are highly non-linear functions often parametrized by a staggering number of weights. Miniaturizing these networks and implementing them in hardware is a direction of research that is fueled by a practical need, and at the same time connects to interesting mathematical problems. For example, by quantizing, or replacing the weights of a neural network with quantized (e.g., binary) counterparts, massive savings in cost, computation time, memory, and power consumption can be attained. Of course, one wishes to attain these savings while preserving the action of the function on domains of interest.
We present data-driven and computationally efficient methods for quantizing the weights of already trained neural networks and we prove that our methods have favorable error guarantees under a variety of assumptions. We also discuss extensions and provide the results of numerical experiments, on large multi-layer networks, to illustrate the performance of our methods. Time permitting, we will also discuss open problems and related areas of research.
|
|
29435
|
Thursday 10/6
2:30 PM
|
Weijie Su, University of Pennsylvania
|
1W-MINDS talk (passcode is the first prime number > 100).
- Weijie Su, University of Pennsylvania
- 1W-MINDS talk (passcode is the first prime number > 100).
- 10/06/2022
- 2:30 PM - 3:30 PM
- Online (virtual meeting)
(Virtual Meeting Link)
- Mark A Iwen (iwenmark@msu.edu)
What Should a Good Deep Neural Network Look Like? Insights from a Layer-Peeled Model and the Law of Equi-Separation
See https://sites.google.com/view/minds-seminar/home
|
|
29437
|
Thursday 10/13
2:30 PM
|
Dustin Mixon, Ohio State University
|
1W-MINDS talk (passcode is the first prime number > 100).
- Dustin Mixon, Ohio State University
- 1W-MINDS talk (passcode is the first prime number > 100).
- 10/13/2022
- 2:30 PM - 3:30 PM
- Online (virtual meeting)
(Virtual Meeting Link)
- Mark A Iwen (iwenmark@msu.edu)
See https://sites.google.com/view/minds-seminar/home
|
|
30456
|
Thursday 10/27
2:30 PM
|
Guanghui Lan, Georgia Institute of Technology
|
1W-MINDS talk (passcode is the first prime number > 100).
- Guanghui Lan, Georgia Institute of Technology
- 1W-MINDS talk (passcode is the first prime number > 100).
- 10/27/2022
- 2:30 PM - 3:30 PM
- Online (virtual meeting)
- Mark A Iwen ()
See https://sites.google.com/view/minds-seminar/home
|
|
30457
|
Thursday 11/3
2:30 PM
|
Alex Townsend , Cornell University
|
1W-MINDS talk (passcode is the first prime number > 100).
- Alex Townsend , Cornell University
- 1W-MINDS talk (passcode is the first prime number > 100).
- 11/03/2022
- 2:30 PM - 3:30 PM
- Online (virtual meeting)
- Mark A Iwen ()
See https://sites.google.com/view/minds-seminar/home
|
|
30458
|
Thursday 11/17
2:30 PM
|
Wei Zhu , University of Massachusetts Amherst
|
1W-MINDS talk (passcode is the first prime number > 100).
- Wei Zhu , University of Massachusetts Amherst
- 1W-MINDS talk (passcode is the first prime number > 100).
- 11/17/2022
- 2:30 PM - 3:30 PM
- Online (virtual meeting)
- Mark A Iwen ()
See https://sites.google.com/view/minds-seminar/home
|
|
30459
|
Thursday 12/8
2:30 PM
|
Rongjie Lai , Rensselaer Polytechnic Institute
|
1W-MINDS talk (passcode is the first prime number > 100).
- Rongjie Lai , Rensselaer Polytechnic Institute
- 1W-MINDS talk (passcode is the first prime number > 100).
- 12/08/2022
- 2:30 PM - 3:30 PM
- Online (virtual meeting)
- Mark A Iwen ()
See https://sites.google.com/view/minds-seminar/home
|
