The Hidden Convexity of Spectral Clustering

Authors

  • James Voss Ohio State University
  • Mikhail Belkin Ohio State University
  • Luis Rademacher Ohio State University

DOI:

https://doi.org/10.1609/aaai.v30i1.10275

Keywords:

spectral clustering, convex maximization, basis recovery

Abstract

In recent years, spectral clustering has become a standard method for data analysis used in a broad range of applications. In this paper we propose a new class of algorithms for multiway spectral clustering based on optimization of a certain "contrast function" over the unit sphere. These algorithms, partly inspired by certain Indepenent Component Analysis techniques, are simple, easy to implement and efficient. Geometrically, the proposed algorithms can be interpreted as hidden basis recovery by means of function optimization. We give a complete characterization of the contrast functions admissible for provable basis recovery. We show how these conditions can be interpreted as a "hidden convexity" of our optimization problem on the sphere; interestingly, we use efficient convex maximization rather than the more common convex minimization. We also show encouraging experimental results on real and simulated data.

Downloads

Published

2016-03-02

How to Cite

Voss, J., Belkin, M., & Rademacher, L. (2016). The Hidden Convexity of Spectral Clustering. Proceedings of the AAAI Conference on Artificial Intelligence, 30(1). https://doi.org/10.1609/aaai.v30i1.10275

Issue

Section

Technical Papers: Machine Learning Methods