Login / Signup
Last Iterate is Slower than Averaged Iterate in Smooth Convex-Concave Saddle Point Problems.
Noah Golowich
Sarath Pattathil
Constantinos Daskalakis
Asuman E. Ozdaglar
Published in:
COLT (2020)
Keyphrases
</>
saddle point
convex concave
variational inequalities
maximum margin
primal dual
penalty function
discrete space
structured prediction
numerical methods
interior point
global constraints
structured output
sensitivity analysis
convex sets
convex hull
markov networks
discrete tomography