NSF funding opportunity: Mathematical & Scientific Foundations of Deep Learning

from: Henry Warchall <hwarchal@nsf.gov>
reply-to: dmsnews <DMSNEWS@listserv.nsf.gov>

A new NSF program solicitation (NSF 20-540) is now available:
NSF-Simons Research Collaborations on the
Mathematical and Scientific Foundations of Deep Learning (MoDL)

Please see https://www.nsf.gov/funding/pgm_summ.jsp?pims_id=505750 for details.

— Letter of Intent (required) Deadline Date: March 20, 2020
— Full Proposal Deadline Date: April 30, 2020

From the Program Solicitation:

Deep learning has been a major driving force in the recent surge of interest in artificial intelligence, both in academia and in industry. In addition to industrial applications, deep learning algorithms are playing important roles in fundamental scientific discoveries. While there has been spectacular empirical success with deep learning, the theoretical understanding of deep learning remains an important emerging research field. To address critical issues, such as the interpretability, reliability, stability, validation, and fairness of algorithms, a convergent effort from mathematicians, statisticians, electrical engineers, and theoretical computer scientists is needed. New mathematical and statistical theories are essential in efficiently dealing with problems, such as approximation, causal inference, convergence, and optimization in high dimensions that are ubiquitous in deep learning.

This program will support up to two new research collaborations. Successful projects are expected to have a cohesive set of goals and a convincing plan that shows that substantial progress will be made in research activities focused on explicit topics involving some of the most challenging questions in the general area of mathematical and scientific foundations of deep learning. Projects are required to bring together theories and approaches from theoretical computer science, electrical engineering, mathematics, and statistics and each project must clearly demonstrate substantial collaborative contributions from members of these four communities. Each project team will conduct training through the research involvement of recent doctoral degree recipients, graduate students, and/or undergraduate students from across this multi-disciplinary spectrum. While the scientific focus must be on the theoretical foundations, relevance to application domains and industry is also important.