We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Mathematics > Differential Geometry

Title: On the geometric and Riemannian structure of the spaces of group equivariant non-expansive operators

Abstract: Group equivariant non-expansive operators have been recently proposed as basic components in topological data analysis and deep learning. In this paper we study some geometric properties of the spaces of group equivariant operators and show how a space $\mathcal{F}$ of group equivariant non-expansive operators can be endowed with the structure of a Riemannian manifold, so making available the use of gradient descent methods for the minimization of cost functions on $\mathcal{F}$. As an application of this approach, we also describe a procedure to select a finite set of representative group equivariant non-expansive operators in the considered manifold.
Comments: 16 pages, 1 figure
Subjects: Differential Geometry (math.DG); Machine Learning (cs.LG)
MSC classes: Primary: 55N31, Secondary: 58D20, 58D30, 62R40, 65D18, 68T09
Cite as: arXiv:2103.02543 [math.DG]
  (or arXiv:2103.02543v1 [math.DG] for this version)

Submission history

From: Patrizio Frosini [view email]
[v1] Wed, 3 Mar 2021 17:29:25 GMT (203kb,D)

Link back to: arXiv, form interface, contact.