Improved Complexity Analysis of the Sinkhorn and Greenkhorn Algorithms for Optimal Transport

The Sinkhorn algorithm is a widely used method for solving the optimal transport problem, and the Greenkhorn algorithm is one of its variants. While there are modified versions of these two algorithms whose computational complexities are O(n2|C|_∞2log n/varepsilon2) to achieve an varepsilon-accuracy, to the best of our knowledge, the existing complexities for the vanilla versions are O(n2|C|_∞3log n/varepsilon3). In this paper we fill this gap and show that the complexities of the vanilla Sinkhorn and Greenkhorn algorithms are indeed O(n2|C|_∞2log n/varepsilon2). The analysis relies on the equicontinuity of the dual variables for the discrete entropic regularized optimal transport problem, which is of independent interest.