35
0

A Surprisingly Simple Method for Distributed Euclidean-Minimum Spanning Tree / Single Linkage Dendrogram Construction from High Dimensional Embeddings via Distance Decomposition

Abstract

We introduce a decomposition method for the distributed calculation of exact Euclidean Minimum Spanning Trees in high dimensions (where sub-quadratic algorithms are not effective), or more generalized geometric-minimum spanning trees of complete graphs, where for each vertex vVv\in V in the graph G=(V,E)G=(V,E) is represented by a vector in vRn\vec{v}\in \mathbb{R}^n, and each for any edge, the the weight of the edge in the graph is given by a symmetric binary `distance' function between the representative vectors w({x,y})=d(x,y)w(\{x,y\}) = d(\vec{x},\vec{y}). This is motivated by the task of clustering high dimensional embeddings produced by neural networks, where low-dimensional algorithms are ineffective; such geometric-minimum spanning trees find applications as a subroutine in the construction of single linkage dendrograms, as the two structures can be converted between each other efficiently.

View on arXiv
Comments on this paper