Abstract:Spectral clustering is one of the most popular and important clustering methods in pattern recognition, machine learning, and data mining. However, its high computational complexity limits it in applications involving truly large-scale datasets. For a clustering problem with n samples, it needs to compute the eigenvectors of the graph Laplacian with O(n3) time complexity. To address this problem, we propose a novel method called anchor-based spectral clustering (ASC) by employing anchor points of data. Specifically, m (m ≪ n) anchor points are selected from the dataset, which can basically maintain the intrinsic (manifold) structure of the original data. Then a mapping matrix between the original data and the anchors is constructed. More importantly, it is proved that this data-anchor mapping matrix essentially preserves the clustering structure of the data. Based on this mapping matrix, it is easy to approximate the spectral embedding of the original data. The proposed method scales linearly relative to the size of the data but with low degradation of the clustering performance. The proposed method, ASC, is compared to the classical spectral clustering and two state-of-the-art accelerating methods, i.e., power iteration clustering and landmark-based spectral clustering, on 10 real-world applications under three evaluation metrics. Experimental results show that ASC is consistently faster than the classical spectral clustering with comparable clustering performance, and at least comparable with or better than the state-of-the-art methods on both effectiveness and efficiency.

Efficient Discrete Clustering with Anchor Graph.

An Anchor-Based Spectral Clustering Method.

Fast Spectral Clustering With Efficient Large Graph Construction

Fast Spectral Clustering via Efficient Multilayer Anchor Graph

Revisiting Fast Spectral Clustering with Anchor Graph

Fast adaptive neighbors clustering via embedded clustering

Attention Non-Negative Spectral Clustering

Deep Spectral Clustering with Constrained Laplacian Rank

Anchor-based Fast Spectral Ensemble Clustering

FGC_SS: Fast Graph Clustering Method by Joint Spectral Embedding and Improved Spectral Rotation

Fast Multi-view Discrete Clustering with Anchor Graphs.

Scalable Fuzzy Clustering With Anchor Graph

Discrete and Balanced Spectral Clustering with Scalability

Fast Clustering by Directly Solving Bipartite Graph Clustering Problem

A Spectral Clustering Algorithm Based on Attribute Fluctuation and Density Peaks Clustering Algorithm

Anchor-free Clustering based on Anchor Graph Factorization

Multiclass Spectral Clustering Based on Discriminant Analysis

Fast Clustering with Anchor Guidance

Local-Global Fuzzy Clustering with Anchor Graph

Spectral clustering with linear embedding: A discrete clustering method for large-scale data

Scalable and Adaptive Spectral Embedding for Attributed Graph Clustering