On the Hardness of Learning Under Symmetries

Explore a 47-minute conference talk presented by Thien Le from the Massachusetts Institute of Technology at IPAM's EnCORE Workshop on Computational vs Statistical Gaps in Learning and Optimization. Delve into the topic "On the hardness of learning under symmetries" as Le discusses the challenges and complexities associated with machine learning algorithms when dealing with symmetrical data structures. Gain insights into the computational and statistical aspects of this problem, and understand how symmetries can impact the learning process. Recorded on February 27, 2024, at the Institute for Pure & Applied Mathematics (IPAM) at UCLA, this presentation offers valuable knowledge for researchers and practitioners in the fields of machine learning, optimization, and computational mathematics.

Thien Le - On the hardness of learning under symmetries - IPAM at UCLA