ML Theory Workshop

The general theme of this workshop is learning theory and the analysis of theoretical problems motivated by modern techniques and applications. While machine learning algorithms are widely adopted in the design of products and services, their use in practice raises a variety of new theoretical and algorithmic challenges, including that of determining guarantees for current methods for training large neural networks, the study of non-convex optimization algorithms, dealing responsibly with key social concerns such as privacy and fairness, the analysis of new distributed learning scenarios and rich interactive learning problems, or the search for enhanced solutions for new scenarios in online learning, active learning, adaptation, time series prediction, and many other stimulating learning problems.

The goal of this workshop is to expose academic researchers to some of the fundamental learning problems appearing in applications and to bring together Google researchers and engineers, as well as researchers from other academic and corporate research laboratories in the NYC area, to discuss solutions and improve our understanding of these technical questions.

The workshop program will include short talks on learning theory topics and theoretical challenges arising in practice, along with open discussions.

Organizers

Co-Organizers