Manifold Learning with Sparse Regularised Optimal Transport

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhang, Stephen, Mordant, Gilles, Matsumoto, Tetsuya, Schiebinger, Geoffrey
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917925521195008
author Zhang, Stephen
Mordant, Gilles
Matsumoto, Tetsuya
Schiebinger, Geoffrey
author_facet Zhang, Stephen
Mordant, Gilles
Matsumoto, Tetsuya
Schiebinger, Geoffrey
contents Manifold learning is a central task in modern statistics and data science. Many datasets (cells, documents, images, molecules) can be represented as point clouds embedded in a high dimensional ambient space, however the degrees of freedom intrinsic to the data are usually far fewer than the number of ambient dimensions. The task of detecting a latent manifold along which the data are embedded is a prerequisite for a wide family of downstream analyses. Real-world datasets are subject to noisy observations and sampling, so that distilling information about the underlying manifold is a major challenge. We propose a method for manifold learning that utilises a symmetric version of optimal transport with a quadratic regularisation that constructs a sparse and adaptive affinity matrix, that can be interpreted as a generalisation of the bistochastic kernel normalisation. We prove that the resulting kernel is consistent with a Laplace-type operator in the continuous limit, establish robustness to heteroskedastic noise and exhibit these results in numerical experiments. We identify a highly efficient computational scheme for computing this optimal transport for discrete data and demonstrate that it outperforms competing methods in a set of examples.
format Preprint
id arxiv_https___arxiv_org_abs_2307_09816
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Manifold Learning with Sparse Regularised Optimal Transport
Zhang, Stephen
Mordant, Gilles
Matsumoto, Tetsuya
Schiebinger, Geoffrey
Machine Learning
Statistics Theory
68T01, 62R30
Manifold learning is a central task in modern statistics and data science. Many datasets (cells, documents, images, molecules) can be represented as point clouds embedded in a high dimensional ambient space, however the degrees of freedom intrinsic to the data are usually far fewer than the number of ambient dimensions. The task of detecting a latent manifold along which the data are embedded is a prerequisite for a wide family of downstream analyses. Real-world datasets are subject to noisy observations and sampling, so that distilling information about the underlying manifold is a major challenge. We propose a method for manifold learning that utilises a symmetric version of optimal transport with a quadratic regularisation that constructs a sparse and adaptive affinity matrix, that can be interpreted as a generalisation of the bistochastic kernel normalisation. We prove that the resulting kernel is consistent with a Laplace-type operator in the continuous limit, establish robustness to heteroskedastic noise and exhibit these results in numerical experiments. We identify a highly efficient computational scheme for computing this optimal transport for discrete data and demonstrate that it outperforms competing methods in a set of examples.
title Manifold Learning with Sparse Regularised Optimal Transport
topic Machine Learning
Statistics Theory
68T01, 62R30
url https://arxiv.org/abs/2307.09816