Conference: Inaugural CAMDA Conference

Machine learning has recently attracted renewed attention following the release of highly versatile large language models such as Chat-GPT and Bard. As concerns about the social and economic impact of technical advances mount, many of machine learning's most successful tools remain poorly understood and their inner working obscure even to their creators. We believe that a healthy and durable society must be built on well-understood, responsible, and interpretable principles. This conference is dedicated to the mathematical foundations of machine learning and its application in analytically tractable settings with the aim of building the tools to better understand the powerful, but inscrutable emerging tools and to discuss interpretable alternative approaches. Texas A&M University is a historical stronghold of approximation theory, which itself underlies learning theory. Indeed, the question 'how well can a function be approximated in general?' arguably precedes the question 'how well can a function be approximated from point values?'. It is an objective of this conference (https://sites.google.com/tamu.edu/camda-conference/) to place rigorous mathematical analysis at the center of future developments in data science in order to guide socially and environmentally responsible progress. Four plenary speakers from Departments of Mathematics, Electrical and Computer Engineering and Computer Science will address an interdisciplinary audience in this effort. An 'open problems' session is planned for the discussion and dissemination of important open problems in this effort, with the specific goal of attracting junior researchers. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.