Mini-Batch Kernel $k$-means

TL;DR

This paper introduces the first mini-batch kernel $k$-means algorithm that significantly accelerates clustering on large datasets with minimal quality loss, supported by theoretical guarantees and extensive experiments.

Contribution

It presents a novel mini-batch kernel $k$-means algorithm with proven convergence and approximation guarantees, enabling scalable kernel clustering.

Findings

Achieves 10-100x speedup over full kernel $k$-means.

Maintains clustering quality with minimal loss.

Provides theoretical analysis with convergence and approximation bounds.

Abstract

We present the first mini-batch kernel $k$-means algorithm, offering an order of magnitude improvement in running time compared to the full batch algorithm. A single iteration of our algorithm takes $\widetilde{O}(kb^2)$ time, significantly faster than the $O(n^2)$ time required by the full batch kernel $k$-means, where $n$ is the dataset size and $b$ is the batch size. Extensive experiments demonstrate that our algorithm consistently achieves a 10-100x speedup with minimal loss in quality, addressing the slow runtime that has limited kernel $k$-means adoption in practice. We further complement these results with a theoretical analysis under an early stopping condition, proving that with a batch size of $\widetilde{\Omega}(\max \{\gamma^{4}, \gamma^{2}\} \cdot \epsilon^{-2})$, the algorithm terminates in $O(\gamma^2/\epsilon)$ iterations with high probability, where $\gamma$ bounds the norm of points in feature space and $\epsilon$ is a termination threshold. Our analysis holds for any reasonable center initialization, and when using $k$-means++ initialization, the algorithm achieves an approximation ratio of $O(\log k)$ in expectation. For normalized kernels, such as Gaussian or Laplacian it holds that $\gamma=1$. Taking $\epsilon = O(1)$ and $b=\Theta(\log n)$, the algorithm terminates in $O(1)$ iterations, with each iteration running in $\widetilde{O}(k)$ time.