Global Convergence of ADMM in Nonconvex Nonsmooth Optimization

Learn about the global convergence properties of the Alternating Direction Method of Multipliers (ADMM) when applied to nonconvex nonsmooth optimization problems in this 45-minute conference talk by Jinshan Zeng presented at ICBS2025 through BIMSA. Explore advanced mathematical concepts and theoretical foundations surrounding ADMM's behavior in challenging optimization landscapes where traditional convexity assumptions do not hold, gaining insights into convergence guarantees and algorithmic performance in nonsmooth settings that commonly arise in machine learning, signal processing, and other computational applications.