Fast Approximate Spectral Clustering for Dynamic Networks

by   Lionel Martin, et al.

Spectral clustering is a widely studied problem, yet its complexity is prohibitive for dynamic graphs of even modest size. We claim that it is possible to reuse information of past cluster assignments to expedite computation. Our approach builds on a recent idea of sidestepping the main bottleneck of spectral clustering, i.e., computing the graph eigenvectors, by using fast Chebyshev graph filtering of random signals. We show that the proposed algorithm achieves clustering assignments with quality approximating that of spectral clustering and that it can yield significant complexity benefits when the graph dynamics are appropriately bounded.


page 1

page 2

page 3

page 4


Spectral Clustering via Graph Filtering: Consistency on the High-Dimensional Stochastic Block Model

Spectral clustering is amongst the most popular methods for community de...

A Dynamic Mode Decomposition Approach for Decentralized Spectral Clustering of Graphs

We propose a novel robust decentralized graph clustering algorithm that ...

Cluster Forests

With inspiration from Random Forests (RF) in the context of classificati...

Simplifying Clustering with Graph Neural Networks

The objective functions used in spectral clustering are usually composed...

Refining Similarity Matrices to Cluster Attributed Networks Accurately

As a result of the recent popularity of social networks and the increase...

Scaling Graph Clustering with Distributed Sketches

The unsupervised learning of community structure, in particular the part...

Spectrally approximating large graphs with smaller graphs

How does coarsening affect the spectrum of a general graph? We provide c...

Please sign up or login with your details

Forgot password? Click here to reset