English

Group Invariant Spectral Embedding

Machine Learning 2026-07-09 v1 Numerical Analysis Statistics Theory

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 MM with symmetries given by a compact Lie group~GG 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/GM/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)\mathrm{SO}(2) or SO(3)\mathrm{SO}(3) symmetry, and show that GG-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.

Cite

@article{arxiv.2607.08987,
  title  = {Group Invariant Spectral Embedding},
  author = {Yeari Vigder and Paulina Hoyos and David Thong and Joakim andén and Joe Kileel and Amit Moscovich},
  journal= {arXiv preprint arXiv:2607.08987},
  year   = {2026}
}