An Improved and Generalised Analysis for Spectral Clustering
By: George Tyler, Luca Zanetti
Potential Business Impact:
Finds hidden groups in connected information.
We revisit the theoretical performances of Spectral Clustering, a classical algorithm for graph partitioning that relies on the eigenvectors of a matrix representation of the graph. Informally, we show that Spectral Clustering works well as long as the smallest eigenvalues appear in groups well separated from the rest of the matrix representation's spectrum. This arises, for example, whenever there exists a hierarchy of clusters at different scales, a regime not captured by previous analyses. Our results are very general and can be applied beyond the traditional graph Laplacian. In particular, we study Hermitian representations of digraphs and show Spectral Clustering can recover partitions where edges between clusters are oriented mostly in the same direction. This has applications in, for example, the analysis of trophic levels in ecological networks. We demonstrate that our results accurately predict the performances of Spectral Clustering on synthetic and real-world data sets.
Similar Papers
Reweighted Spectral Partitioning Works: Bounds for Special Graph Classes
Data Structures and Algorithms
Finds better ways to cut networks apart.
Reweighted Spectral Partitioning Works: Bounds for Special Graph Classes
Data Structures and Algorithms
Finds better ways to split computer networks.
Structure-Aware Spectral Sparsification via Uniform Edge Sampling
Machine Learning (CS)
Makes computer grouping faster on huge networks.