Salvato in:
Dettagli Bibliografici
Autori principali: Rana, Muhammad, Tasissa, Abiy, Cai, HanQin, Gavriyelov, Yakov, Hamm, Keaton
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2509.19250
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866911695332442112
author Rana, Muhammad
Tasissa, Abiy
Cai, HanQin
Gavriyelov, Yakov
Hamm, Keaton
author_facet Rana, Muhammad
Tasissa, Abiy
Cai, HanQin
Gavriyelov, Yakov
Hamm, Keaton
contents This paper proposes two algorithms for estimating square Wasserstein distance matrices from a small number of entries. These matrices are used to compute manifold learning embeddings like multidimensional scaling (MDS) or Isomap, but contrary to Euclidean distance matrices, are extremely costly to compute. We analyze matrix completion from upper triangular samples and Nyström completion in which $\mathcal{O}(d\log(d))$ columns of the distance matrices are computed where $d$ is the desired embedding dimension, prove stability of MDS under Nyström completion, and show that it can outperform matrix completion for a fixed budget of sample distances. Finally, we show that classification of the OrganCMNIST dataset from the MedMNIST benchmark is stable on data embedded from the Nyström estimation of the distance matrix even when only 10\% of the columns are computed.
format Preprint
id arxiv_https___arxiv_org_abs_2509_19250
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Recovering Wasserstein Distance Matrices from Few Measurements
Rana, Muhammad
Tasissa, Abiy
Cai, HanQin
Gavriyelov, Yakov
Hamm, Keaton
Machine Learning
This paper proposes two algorithms for estimating square Wasserstein distance matrices from a small number of entries. These matrices are used to compute manifold learning embeddings like multidimensional scaling (MDS) or Isomap, but contrary to Euclidean distance matrices, are extremely costly to compute. We analyze matrix completion from upper triangular samples and Nyström completion in which $\mathcal{O}(d\log(d))$ columns of the distance matrices are computed where $d$ is the desired embedding dimension, prove stability of MDS under Nyström completion, and show that it can outperform matrix completion for a fixed budget of sample distances. Finally, we show that classification of the OrganCMNIST dataset from the MedMNIST benchmark is stable on data embedded from the Nyström estimation of the distance matrix even when only 10\% of the columns are computed.
title Recovering Wasserstein Distance Matrices from Few Measurements
topic Machine Learning
url https://arxiv.org/abs/2509.19250