A proof blueprint establishes robust O(1/k) rates for entropic Bregman projections that scale linearly in the inverse regularization strength, instantiated as a new flow-Sinkhorn method for graph W1 with O(p diameter^3 / ε^4) complexity.
Better and simpler error analysis of the sinkhorn–knopp algorithm for matrix scaling.Mathematical Programming, 188(1):395–407
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.OC 1years
2026 1verdicts
ACCEPT 1representative citing papers
citing papers explorer
-
Robust Sublinear Convergence Rates for Iterative Bregman Projections
A proof blueprint establishes robust O(1/k) rates for entropic Bregman projections that scale linearly in the inverse regularization strength, instantiated as a new flow-Sinkhorn method for graph W1 with O(p diameter^3 / ε^4) complexity.