English(EN) Reliable Modeling of Distribution Shifts via Displacement-Reshaped Optimal Transport

新的 ReshapeOT 方法改进了用于建模分布偏移的最优传输

作者 PulseAugur 编辑部 · [2 个来源] · 2026-05-06 14:24

研究人员推出了一种新颖的分布偏移建模方法——位移重塑最优传输 (ReshapeOT)。该技术通过纳入观察到的样本位移来增强最优传输中使用的地面度量。通过用源自位移矩的马氏距离替换标准欧氏距离，ReshapeOT 指导传输解决方案更好地反映数据的实际变化。 AI

影响引入了一种更可靠地建模分布偏移的新方法，有可能提高各种机器学习应用的性能。

排序理由该集群包含一篇 arXiv 预印本，详细介绍了一种用于建模分布偏移的新方法。

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

arXiv cs.LG TIER_1 English(EN) · Philip Naumann, Jacob Kauffmann, Klaus-Robert M\"uller, Gr\'egoire Montavon · 2026-05-07 04:00

Reliable Modeling of Distribution Shifts via Displacement-Reshaped Optimal Transport

arXiv:2605.04965v1 Announce Type: new Abstract: Optimal transport (OT) is a central framework for modeling distribution shifts. Because OT compares distributions directly in input space, a well-designed ground metric between observations is essential to ensure that the optimizer …
arXiv cs.AI TIER_1 English(EN) · Grégoire Montavon · 2026-05-06 14:24

Reliable Modeling of Distribution Shifts via Displacement-Reshaped Optimal Transport

Optimal transport (OT) is a central framework for modeling distribution shifts. Because OT compares distributions directly in input space, a well-designed ground metric between observations is essential to ensure that the optimizer does not violate the true geometry of change. We…