Learning Spectral Clustering - WebHome.pdfVIP

Learning Spectral Clustering Francis R. Bach Michael I. Jordan Computer Science Computer Science and Statistics University of California University of California Berkeley, CA 94720 Berkeley, CA 94720 fbach@ jordan@ Abstract Spectral clustering refers to a class of techniques which rely on the eigen- structure of a similarity matrix to partition points into disjoint clusters with points in the same cluster having high similarity and points in dif- ferent clusters having low similarity. In this paper, we derive a new cost function for spectral clustering based on a measure of error between a given partition and a solution of the spectral relaxation of a minimum normalized cut problem. Minimizing this cost function with respect to the partition leads to a new spectral clustering algorithm. Minimizing with respect to the similarity matrix leads to an algorithm for learning the similarity matrix. We develop a tractable approximation of our cost function that is based on the power method of computing eigenvectors. 1 Introduction Spectral clustering has many applications in machine learning, exploratory data analysis, computer vision and speech processing. Most techniques explicitly or implicitly assume a metric or a similarity structure over the space of configurations, which is then used by clustering algorithms. The success of such algorithms depends heavily on the choice of the metric, but this choice is generally not treated as part of the learning problem. Thus, time-consuming manual feature selection and weighting is often a necessary precursor to the use of spectral methods. Several recent papers have conside