AI RESEARCH
On the Role of Batch Size in Stochastic Conditional Gradient Methods
arXiv CS.LG
•
ArXi:2603.21191v1 Announce Type: new We study the role of batch size in stochastic conditional gradient methods under a $\mu$-Kurdyka-{\L}ojasiewicz ($\mu$-KL) condition. Focusing on momentum-based stochastic conditional gradient algorithms (e.g., Scion), we derive a new analysis that explicitly captures the interaction between stepsize, batch size, and stochastic noise.