AI RESEARCH
On the Relationship Between CoCoA and ADMM for Distributed Empirical Risk Minimization
arXiv CS.LG
•
ArXi:2502.00470v3 Announce Type: replace-cross Distributed empirical risk minimization (ERM) is often studied through two influential yet seemingly separate families of methods: CoCoA-type algorithms, derived from distributed dual coordinate ascent, and ADMM-type algorithms, derived from consensus and proximal splitting. In this paper, we investigate the connection of the two types of algorithms from a unified primal-dual perspective.