Algorithmic stability: mathematical foundations for the modern era
May 12 to May 16, 2025
at the
American Institute of Mathematics,
Pasadena, California
organized by
Rina Foygel Barber and Jake Soloff
Original Announcement
This workshop will be devoted to building a foundational understanding of algorithmic stability, and developing rigorous tools for measuring stability that can characterize the behavior of machine learning algorithms. We aim to bring together researchers across a broad range of fields to develop a unified theoretical foundation for algorithmic stability.
The main topics for the workshop are
-
Building an understanding of the relationships between, and strengths and limitations of, different frameworks and definitions of stability.
-
The role of algorithmic stability in overparameterized learning models, both in theory and in practice.
-
Statistical tools for quantifying stability and stabilizing learning algorithms.
-
Building a shared theoretical framework to bridge the gap between differential privacy and stability.
Material from the workshop
A list of participants.
The workshop schedule.
