Focus Period lund 2026
PhD Student
Saarland University & CISPA Helmholtz Center for Information Security (Germany)
Yuan Gao is a PhD student at Saarland University and CISPA Helmholtz Center for Information Security. He is a member of the Machine Learning and Optimization group at CISPA, working with Sebastian Stich. He is broadly interested in the theory of optimization, and in particular gradient methods and their applications in Machine Learning.
Presenting: Revisiting mirror descent for incremental gradient.
In this talk, I will revisit the incremental gradient method and analyse some dual averaging variants of the mirror descent method. I will discuss the implementation and convergence of the methodand identify some of their potential further applications in more general settings.