Regularized Non-negative Spectral Embedding for Clustering

TL;DR

This paper introduces RNSE, an end-to-end spectral clustering method that jointly learns similarity and clustering labels, improving performance over traditional multi-stage spectral clustering methods.

Contribution

RNSE is a novel single-stage spectral clustering approach that adaptively learns the similarity matrix and enforces non-negativity for direct clustering from data.

Findings

RNSE outperforms state-of-the-art methods on synthetic datasets.

RNSE achieves superior results on real-world datasets.

The optimization techniques effectively solve complex problems in RNSE.

Abstract

Spectral Clustering is a popular technique to split data points into groups, especially for complex datasets. The algorithms in the Spectral Clustering family typically consist of multiple separate stages (such as similarity matrix construction, low-dimensional embedding, and K-Means clustering as post processing), which may lead to sub-optimal results because of the possible mismatch between different stages. In this paper, we propose an end-to-end single-stage learning method to clustering called Regularized Non-negative Spectral Embedding (RNSE) which extends Spectral Clustering with the adaptive learning of similarity matrix and meanwhile utilizes non-negative constraints to facilitate one-step clustering (directly from data points to clustering labels). Two well-founded methods, successive alternating projection and strategic multiplicative update, are employed to work out the quite challenging optimization problems in RNSE. Extensive experiments on both synthetic and real-world datasets demonstrate RNSE superior clustering performance to some state-of-the-art competitors.