LMS Invited Lectures on the Mathematics of Deep Learning
| Created: | 2022-03-08 14:50 |
|---|---|
| Institution: | Isaac Newton Institute for Mathematical Sciences |
| Description: | Monday 28th February 2022 to Friday 4th March 2022
Background This workshop featured an introductory lecture series across the week, by Professor Gitta Kutyniok (LMU, Munich) on the mathematics of deep learning. Four accompanying lectures have also been given. Neural networks were originally introduced in 1943 by McCulloch and Pitts as an approach to develop learning algorithms by mimicking the human brain. The key goal at that time was the introduction of a theory of artificial intelligence. However, the limited amount of data and the lack of high-performance computers made the training of deep neural networks, (networks with many layers), unfeasible. Today, massive amounts of training data are available complemented by a tremendously increased computing power, allowing for the first time the application of deep learning algorithms. It is for this reason that deep neural networks have recently seen an impressive comeback. Spectacular applications of deep learning are AlphaGo, which for the first time enabled a computer to beat the top world players in the game Go - a game by far more complex than chess-, or the speech recognition systems available on each smartphone these days. We currently witness how algorithms based on deep neural networks are used in numerous aspects of the public sector such as being used for pre-screening job applications or revolutionizing the healthcare industry. In fact, the U.S. Food and Drug Administration (FDA) has already approved the marketing of the first medical device for detecting diabetic retinopathy which is based on such methodologies. Aims and Objectives The workshop aimed to provide a mathematical foundation of deep learning by an introduction to the main mathematical questions and concepts of deep neural networks and their training within two realms: Theoretical foundations of deep learning independent of a particular application Theoretical analysis of the potential and the limitations of deep learning for mathematical methodologies, in particular, for inverse problems and partial differential equations. The programme featured 2 lectures each day as well as the opportunity for discussion and networking. A poster session also took place. |
| Website: | https://gateway.newton.ac.uk/event/tgm109 |
Media items
This collection contains 14 media items.
Media items
Benign Overfitting in Linear and Nonlinear Settings
Peter Bartlett (University of California, Berkeley)
3 March 2022 – 15:30 to 16:30
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
Deep Neural Networks: Analysing the Training Algorithm
Gitta Kutyniok (Ludwig-Maximilians-Universität München)
1 March 2022 – 10:00 to 11:00
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
Deep Neural Networks: From Approximation to Expressivity
Gitta Kutyniok (Ludwig-Maximilians-Universität München)
28 February 2022 – 16:00 to 17:00
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
Deep Neural Networks: Opening the Black Box via Explainability Methods
Gitta Kutyniok (Ludwig-Maximilians-Universität München)
2 March 2022 – 10:00 to 11:00
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
Deep Neural Networks: The Mystery of Generalisation
Gitta Kutyniok (Ludwig-Maximilians-Universität München)
1 March 2022 – 14:00 to 15:00
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
Deep Neural Networks: Towards Robustness
Gitta Kutyniok (Ludwig-Maximilians-Universität München)
2 March 2022 – 11:30 to 12:30
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
Introduction to Deep Neural Networks
Gitta Kutyniok (Ludwig-Maximilians-Universität München)
28 February 2022 – 14:15 to 15:15
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
Inverse Problems meet Deep Learning: Optimal Hybrid Methods
Gitta Kutyniok (Ludwig-Maximilians-Universität München)
3 March 2022 – 14:00 to 15:00
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
Limitations of Deep Neural Networks
Gitta Kutyniok (Ludwig-Maximilians-Universität München)
3 March 2022 – 10:00 to 11:00
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
Machine Learning for the Sciences: Towards Understanding
Klaus-Robert Müller (Technische Universität Berlin)
3 March 2022 – 11:30 to 12:30
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
Mathematical Foundations of Deep Learning: Potential, Limitations, and Future Directions
Gitta Kutyniok (Ludwig-Maximilians-Universität München)
4 March 2022 – 14:00 to 15:00
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Wed 9 Mar 2022
Mathematics of Deep Learning: Some Open Problems
Weinan E (Princeton University)
4 March 2022 – 10:00 to 11:00
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
Partial Differential Equations meet Deep Learning: Beating the Curse of Dimensionality
Gitta Kutyniok (Ludwig-Maximilians-Universität München)
4 March 2022 – 11:30 to 12:30
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022
The Role of Linear Layers in Nonlinear Interpolating Networks
Rebecca Willett (University of Chicago)
2 March 2022 – 16:00 to 17:00
Collection: LMS Invited Lectures on the Mathematics of Deep Learning
Institution: Isaac Newton Institute for Mathematical Sciences
Created: Tue 8 Mar 2022