Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox

Picture for Sarath Pattathil

Tight last-iterate convergence rates for no-regret learning in multi-player games


Oct 26, 2020
Noah Golowich, Sarath Pattathil, Constantinos Daskalakis

* To appear at NeurIPS 2020. 41 pages 

  Access Paper or Ask Questions

An Optimal Multistage Stochastic Gradient Method for Minimax Problems


Feb 13, 2020
Alireza Fallah, Asuman Ozdaglar, Sarath Pattathil


  Access Paper or Ask Questions

Last Iterate is Slower than Averaged Iterate in Smooth Convex-Concave Saddle Point Problems


Jan 31, 2020
Noah Golowich, Sarath Pattathil, Constantinos Daskalakis, Asuman Ozdaglar

* 25 pages 

  Access Paper or Ask Questions

A Decentralized Proximal Point-type Method for Saddle Point Problems


Oct 31, 2019
Weijie Liu, Aryan Mokhtari, Asuman Ozdaglar, Sarath Pattathil, Zebang Shen, Nenggan Zheng

* 18 pages 

  Access Paper or Ask Questions

Proximal Point Approximations Achieving a Convergence Rate of $\mathcal{O}(1/k)$ for Smooth Convex-Concave Saddle Point Problems: Optimistic Gradient and Extra-gradient Methods


Jun 03, 2019
Aryan Mokhtari, Asuman Ozdaglar, Sarath Pattathil

* 15 pages 

  Access Paper or Ask Questions

A Unified Analysis of Extra-gradient and Optimistic Gradient Methods for Saddle Point Problems: Proximal Point Approach


Jan 24, 2019
Aryan Mokhtari, Asuman Ozdaglar, Sarath Pattathil

* 31 pages, 3 figures 

  Access Paper or Ask Questions