
TL;DR
This paper investigates nonlinear spectral gaps in metric spaces, establishing conditions for embeddability and nonembeddability, and provides new bounds on the distortion of specific graph families.
Contribution
It offers a geometric characterization of metric space pairs based on nonlinear spectral gaps and derives new embeddability and nonembeddability results.
Findings
Established a linear characterization for metric space pairs with spectral gap inequalities.
Proved existence of embeddings from $L_p$ to $L_2$ with controlled distortion for $p>2$.
Improved lower bounds on the $L_p$ distortion of Ramanujan graphs.
Abstract
Let be an by symmetric stochastic matrix. For and a metric space , let be the infimum over those for which every satisfy Thus measures the magnitude of the {\em nonlinear spectral gap} of the matrix with respect to the kernel . We study pairs of metric spaces and for which there exists such that for every symmetric stochastic with . When is linear a complete geometric characterization is obtained. Our estimates on nonlinear spectral gaps yield new…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
