pith. sign in

Accelerated Sinkhorn Algorithms for Partial Optimal Transport

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it
abstract

Partial Optimal Transport (POT) addresses the problem of transporting only a fraction of the total mass between two distributions, making it suitable when marginals have unequal size or contain outliers. While Sinkhorn-based methods are widely used, their complexity bounds for POT remain suboptimal and can limit scalability. We introduce Accelerated Sinkhorn for POT (ASPOT), which integrates alternating minimization with Nesterov-style acceleration in the POT setting, yielding a complexity of $\mathcal{O}(n^{7/3}\varepsilon^{-5/3})$. We also show that an informed choice of the entropic parameter $\gamma$ improves rates for the classical Sinkhorn method. Experiments on real-world applications validate our theories and demonstrate the favorable performance of our proposed methods.

fields

cs.LG 1

years

2026 1

verdicts

UNVERDICTED 1

representative citing papers

citing papers explorer

Showing 1 of 1 citing paper.

  • Accelerated Sinkhorn Algorithms for Partial Optimal Transport cs.LG · 2026-01-23 · unverdicted · none · ref 7 · internal anchor

    ASPOT achieves O(n^{7/3} ε^{-5/3}) complexity for partial optimal transport via Nesterov-accelerated alternating minimization, plus an improved gamma choice for classical Sinkhorn.