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.

Revisiting Fast Spectral Clustering with Anchor Graph

Fast Spectral Clustering With Efficient Large Graph Construction

Efficient Discrete Clustering with Anchor Graph.

An Anchor-Based Spectral Clustering Method.

Fast Fuzzy Clustering Based on Anchor Graph

Fast Spectral Clustering With Anchor Graph for Large Hyperspectral Images.

Anchor-based Fast Spectral Ensemble Clustering

Fast Spectral Clustering via Efficient Multilayer Anchor Graph

A fast adaptive nearest neighbor clustering method based on structured anchor graph

Scalable Fuzzy Clustering With Anchor Graph

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

Fast Multi-view Discrete Clustering with Anchor Graphs.

Fast Clustering with Anchor Guidance

Fast spectral clustering with self-weighted features

Fast Density-Weighted Low-Rank Approximation Spectral Clustering

Fast Self-Supervised Clustering With Anchor Graph

Attention Non-Negative Spectral Clustering

Fast Spectral Clustering Learning with Hierarchical Bipartite Graph for Large-Scale Data

Fast local representation learning with adaptive anchor graph

Local-Global Fuzzy Clustering with Anchor Graph

Fast adaptive neighbors clustering via embedded clustering