Scalable Constrained Clustering: A Generalized Spectral Method
Social and Information Networks
2016-01-20 v1
Abstract
We present a simple spectral approach to the well-studied constrained clustering problem. It captures constrained clustering as a generalized eigenvalue problem with graph Laplacians. The algorithm works in nearly-linear time and provides concrete guarantees for the quality of the clusters, at least for the case of 2-way partitioning. In practice this translates to a very fast implementation that consistently outperforms existing spectral approaches both in speed and quality.
Cite
@article{arxiv.1601.04746,
title = {Scalable Constrained Clustering: A Generalized Spectral Method},
author = {Mihai Cucuringu and Ioannis Koutis and Sanjay Chawla and Gary Miller and Richard Peng},
journal= {arXiv preprint arXiv:1601.04746},
year = {2016}
}
Comments
accepted to appear in AISTATS 2016. arXiv admin note: text overlap with arXiv:1504.00653