Coreset Spectral Clustering
This work addresses scalability issues in graph clustering for large datasets, offering a hybrid approach that is incremental but with specific performance gains.
The paper tackles the problem of scaling spectral clustering to large graphs with many clusters by combining coresets and kernel k-means, achieving an O(α)-approximation guarantee and improving runtime from O(nk) to O(n·min{k, d_avg}) for sparse kernels.
Coresets have become an invaluable tool for solving $k$-means and kernel $k$-means clustering problems on large datasets with small numbers of clusters. On the other hand, spectral clustering works well on sparse graphs and has recently been extended to scale efficiently to large numbers of clusters. We exploit the connection between kernel $k$-means and the normalised cut problem to combine the benefits of both. Our main result is a coreset spectral clustering algorithm for graphs that clusters a coreset graph to infer a good labelling of the original graph. We prove that an $α$-approximation for the normalised cut problem on the coreset graph is an $O(α)$-approximation on the original. We also improve the running time of the state-of-the-art coreset algorithm for kernel $k$-means on sparse kernels, from $\tilde{O}(nk)$ to $\tilde{O}(n\cdot \min \{k, d_{avg}\})$, where $d_{avg}$ is the average number of non-zero entries in each row of the $n\times n$ kernel matrix. Our experiments confirm our coreset algorithm is asymptotically faster on large real-world graphs with many clusters, and show that our clustering algorithm overcomes the main challenge faced by coreset kernel $k$-means on sparse kernels which is getting stuck in local optima.