Project Grant N000142612074

This National Science Foundation (NSF) Computer and Information Science and Engineering (CISE) program award of $508,848 to the University of California, Berkeley aims to advance the theoretical understanding of deep learning algorithms and architectures. The project has three main thrusts: 1) analyzing the algorithmic regularization effects of deep learning approaches to improve accuracy, 2) theoretically comparing the performance of neural networks trained with stochastic gradient descent...

This four-year $800,000 Project Grant from the National Science Foundation's (NSF) Computer and Information Science and Engineering program (CFDA 47.070) will support research at the University of California, Los Angeles (UCLA) to develop theoretical tools for understanding deep neural networks (DNNs). The research aims to identify common themes in how artificial and biological systems like the human brain learn. It will investigate the hypothesis that DNNs succeed when learning tasks exhibit...

This three-year Project Grant from the National Science Foundation's Division of Computing and Communication Foundations under the Computer and Information Science and Engineering program will support the development of new optimization approaches for machine learning problems. The $600,000 award to the University of Wisconsin-Madison beginning October 1, 2022 will advance optimization algorithms and analysis techniques for convex-concave minimax problems incorporating sparsity or regularity....

This two-year, $250,000 Project Grant from the National Science Foundation's Computer and Information Science and Engineering program aims to develop techniques for improving the interpretability and robustness of deep neural networks. Specifically, the University of California, Santa Barbara will apply ideas from communication theory and neuroscience to actively shape the features extracted by individual layers of neural networks in addition to end-to-end training. By learning "matched...

This three-year National Science Foundation project grant of $301,244 will fund research into improving reinforcement learning methods through analyzing their connection to gradient descent. The grant was awarded by the NSF Division of Computing and Communication Foundations under the Computer and Information Science and Engineering program. The grantee, Trustees of Boston University, will explore a novel approach to analyzing and designing convergent reinforcement learning algorithms based on a...

The National Science Foundation (NSF), through its Computer and Information Science and Engineering program (CFDA 47.070), has awarded a $400,000 project grant to The Ohio State University to conduct research on the theoretical principles underlying the success of deep learning models. The 3-year award, effective July 1, 2023, will focus on three main thrusts: Developing a unified mathematical framework to analyze the convergence and neural collapse properties of overparameterized deep...

This three-year project grant from the National Science Foundation's Mathematical and Physical Sciences program (CFDA 47.049) provides $200,000 to support research at the University of Utah developing new deep learning algorithms for sequential and graph data motivated by differential equations theory. The grant aims to advance machine learning methodology for ubiquitous problems involving sequentially observed data from multiple agents, such as pandemic spread modeling, cooperative robotics...

This federal Project Grant award from the National Science Foundation's Mathematical and Physical Sciences program (CFDA 47.049) provides $175,000 to the University of California, Berkeley to develop advanced nonparametric regression techniques for analyzing complex data. The key products of this research include new shape-constrained methods for multi-index convex regression to address the "curse of dimensionality," as well as the systematic study of total generalized variation...

This National Science Foundation (NSF) Project Grant award provides $399,994 to The Johns Hopkins University from July 1, 2023 to June 30, 2026. The grant is funded under the NSF's Computer and Information Science and Engineering program (CFDA 47.070), which supports research and education in computing, communications, and information science. The project aims to develop a rigorous mathematical theory to explain the phenomenon of "neural collapse" in deep learning models, and use...

This Project Grant award from the National Science Foundation's (NSF) Computer and Information Science and Engineering (CISE) program, with CFDA number 47.070, provides $266,000.00 to The Leland Stanford Junior University (Stanford University) to advance research on "Advancing Large Language Model Unlearning: Foundations and Applications." The key objectives of this 4-year project are to develop new algorithmic frameworks to enhance the effectiveness, robustness, and efficiency of...