Abstract:Minimum spanning trees and forests are powerful sparsification techniques that remove cycles from weighted graphs to minimize total edge weight while preserving node connectivity. They have applications in computer science, network science, and graph theory. Despite their utility and ubiquity, they have several limitations, including that they are only defined for undirected networks, they significantly alter dynamics on networks, and they do not generally preserve important network features such as shortest distances, shortest path distribution, and community structure. In contrast, distance backbones, which are subgraphs formed by all edges that obey a generalized triangle inequality, are well defined in both directed and undirected graphs and preserve those and other important network features. The backbone of a graph is defined with respect to a specified path-length operator that aggregates weights along a path to define its length, thereby associating a cost to indirect connections. The backbone is the union of all shortest paths between each pair of nodes according to the specified operator. One such operator, the max function, computes the length of a path as the largest weight of the edges that compose it (a weakest link criterion). It is the only operator that yields an algebraic structure for computing shortest paths that is consistent with De Morgan's laws. Applying this operator yields the ultrametric backbone of a graph in that (semi-triangular) edges whose weights are larger than the length of an indirect path connecting the same nodes (i.e., those that break the generalized triangle inequality based on max as a path-length operator) are removed. We show that the ultrametric backbone is the union of all minimum spanning forests in undirected graphs and provides a new generalization of minimum spanning trees to directed graphs.

The Matrix of Maximum Out Forests of a Digraph and Its Applications

Spin-dependent hole quantum transport in Aharonov-Bohm ring structure: possible schemes for spin filter

Digraph Branchings and Matrix Determinants

Matrices of forests, analysis of networks, and ranking problems

Forest matrices around the Laplacian matrix

Maximum induced forests in random graphs

Random minimum spanning tree and dense graph limits

A Row-wise Algorithm for Graph Realization

Maximum induced trees and forests of bounded degree in random graphs

Complexity Dichotomies for the Maximum Weighted Digraph Partition Problem

Finding Subgraphs with Maximum Total Density and Limited Overlap in Weighted Hypergraphs

Bottleneck matrices of maximal outerplanar graphs with isomorphic underlying trees

Into the Woods: A graph theoretic proof of the Dodgson/Muir identity

Bilinear matrix equation characterizes Laplacian and distance matrices of weighted trees

A Maximum Linear Arrangement Problem on Directed Graphs

A Graph Bottleneck Inequality

The ultrametric backbone is the union of all minimum spanning forests

A $\frac{4}{3}$-Approximation for the Maximum Leaf Spanning Arborescence Problem in DAGs

Learning Spanning Forests Optimally using CUT Queries in Weighted Undirected Graphs

Leafy Spanning Arborescences in DAGs

Reversible Mcmc on Markov Equivalence Classes of Sparse Directed Acyclic Graphs