Fused Gromov-Wasserstein distance for structured objects: theoretical foundations and mathematical properties · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/1811.02834/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTFused Gromov-Wasserstein distance for structured objects: theoretical foundations and mathematical properties — mentioned by bionhowardhttps://news.ycombinator.com/item?id=22395703urn:hackerpapers:mention:1811.02834:22395703Sun, 23 Feb 2020 08:03:39 GMT<p>is the loss you&#x27;re talking about like fused gromov-wasserstein distance? we hit permutation invariance issues like what you&#x27;re talking about in some atomistic simulations because the atoms need to be permutable if you want to use the same model for chemistry as protein folding/docking, and the FGW algo from e.g. https://arxiv.org/pdf/1811.02834.pdf https://tvayer.github.io/materials/Titouan_Marseille_2019.pd... relaxes the invariance issue by adding a feature distance to the euclidean distance. higher order distance matrices are a neat trick, but blow up VRAM past 10-50k atoms, but if you did…</p><p><a href="https://hackerpaper.kaipfister.de/p/1811.02834/">Paper and discussion history</a>. A mention is not an endorsement.</p>stat.ML