Curvature, diameter and signs of graphs
Wei Chen, Shiping Liu

TL;DR
This paper establishes eigenvalue-diameter bounds for non-negatively curved signed graphs, leading to volume estimates and Li-Yau inequalities, using gradient estimates and nodal domain techniques, with extensions to nonlinear Laplacians.
Contribution
It introduces a Li-Yau type eigenvalue-diameter estimate for signed graphs and develops new methods involving nodal domain walks, extending results to nonlinear Laplacians.
Findings
Eigenvalues are bounded below by 1/D^2 for non-negatively curved signed graphs.
Volume estimates relate frustration index and diameter.
Two-sided Li-Yau estimates are obtained for triangle-free graphs.
Abstract
We prove a Li-Yau type eigenvalue-diameter estimate for signed graphs. That is, the nonzero eigenvalues of the Laplacian of a non-negatively curved signed graph are lower bounded by up to a constant, where stands for the diameter. This leads to several interesting applications, including a volume estimate for non-negatively curved signed graphs in terms of frustration index and diameter, and a two-sided Li-Yau estimate for triangle-free graphs. Our proof is built upon a combination of Chung-Lin-Yau type gradient estimate and a new trick involving strong nodal domain walks of signed graphs. We further discuss extensions of part of our results to nonlinear Laplacians on signed graphs.
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.
Taxonomy
TopicsHistory and Theory of Mathematics · Mathematics and Applications
