Group Invariant Spectral Embedding
Abstract
Spectral embedding methods are widely used for dimensionality reduction and clustering of high-dimensional datasets with intrinsic low-dimensional structures. Although many datasets of practical interest exhibit invariance under symmetries such as rotations, standard spectral embedding methods do not account for this, treating symmetry-related data points as unrelated. Our approach to this problem is to incorporate the symmetries directly into the affinity kernels used for spectral embedding. We analyze the case of a Riemannian data manifold M with symmetries given by a compact Lie group~G and prove that, under suitable conditions, graph Laplacians constructed from three types of invariant kernels converge pointwise to explicit second-order differential operators on the quotient space M/G. Our analysis implies improved convergence rates, as the effective dimension drops according to the dimension of the group. We validate our approach on datasets with SO(2) or SO(3) symmetry, and show that G-invariant spectral embedding recovers the intrinsic geometry of the data, in contrast to standard spectral embedding, which fails to do so even in the limit of infinite data.
Related
- Quandles from group actions and a Cayley-type embedding theorem
- A Lipschitz refinement of the Bebutov-Kakutani dynamical embedding theorem for connected Lie group actions
- Generic Sharp Shift Embedding for Amenable Group Actions with the Uniform Rokhlin Property and Comparison
- Spectral embedding through weak* limit of finite-dimensional approximations
- Enhancing Spectral Embedding through Robust and Flexible Knowledge Transfer in Electronic Health Records
BibTeX
@misc{vigder2026group,
title = {{Group Invariant Spectral Embedding}},
author = {Yeari Vigder and Paulina Hoyos and David Thong and Joakim andén and Joe Kileel and Amit Moscovich},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2607.08987},
eprint = {2607.08987},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2607.08987},
}
RIS
TY - UNPB
TI - Group Invariant Spectral Embedding
AU - Yeari Vigder
AU - Paulina Hoyos
AU - David Thong
AU - Joakim andén
AU - Joe Kileel
AU - Amit Moscovich
PY - 2026
DA - 2026/10/08
PB - arXiv
DO - 10.48550/arxiv.2607.08987
UR - https://arxiv.org/abs/2607.08987
AB - Spectral embedding methods are widely used for dimensionality reduction and clustering of high-dimensional datasets with intrinsic low-dimensional structures. Although many datasets of practical interest exhibit invariance under symmetries such as rotations, standard spectral embedding methods do not account for this, treating symmetry-related data points as unrelated. Our approach to this problem is to incorporate the symmetries directly into the affinity kernels used for spectral embedding. We analyze the case of a Riemannian data manifold M with symmetries given by a compact Lie group~G and prove that, under suitable conditions, graph Laplacians constructed from three types of invariant kernels converge pointwise to explicit second-order differential operators on the quotient space M/G. Our analysis implies improved convergence rates, as the effective dimension drops according to the dimension of the group. We validate our approach on datasets with SO(2) or SO(3) symmetry, and show that G-invariant spectral embedding recovers the intrinsic geometry of the data, in contrast to standard spectral embedding, which fails to do so even in the limit of infinite data.
ER -