Variance Reduction for Matrix Games
Yair Carmon, Yujia Jin, Aaron Sidford and Kevin Tian
Stanford University
{yairc,yujiajin,sidford,kjtian}@stanford.edu
Abstract
We present a randomized primal-dual algorithm thatpsolves the problem
minx maxy y > Ax to additive error in time nnz(A) + nnz(A)n/, for ma-
trix A with larger dimension n and nnz(A) nonzero
p entries. This improves the best
...


雷达卡



京公网安备 11010802022788号







