English

On Optimal Coreset Construction for Euclidean $(k,z)$-Clustering

Data Structures and Algorithms 2024-03-14 v6 Computational Geometry

Abstract

Constructing small-sized coresets for various clustering problems in different metric spaces has attracted significant attention for the past decade. A central problem in the coreset literature is to understand what is the best possible coreset size for (k,z)(k,z)-clustering in Euclidean space. While there has been significant progress in the problem, there is still a gap between the state-of-the-art upper and lower bounds. For instance, the best known upper bound for kk-means (z=2z=2) is min{O(k3/2ε2),O(kε4)}\min \{O(k^{3/2} \varepsilon^{-2}),O(k \varepsilon^{-4})\} [1,2], while the best known lower bound is Ω(kε2)\Omega(k\varepsilon^{-2}) [1]. In this paper, we make significant progress on both upper and lower bounds. For a large range of parameters (i.e., ε,k\varepsilon, k), we have a complete understanding of the optimal coreset size. In particular, we obtain the following results: (1) We present a new coreset lower bound Ω(kεz2)\Omega(k \varepsilon^{-z-2}) for Euclidean (k,z)(k,z)-clustering when εΩ(k1/(z+2))\varepsilon \geq \Omega(k^{-1/(z+2)}). In view of the prior upper bound O~z(kεz2)\tilde{O}_z(k \varepsilon^{-z-2}) [1], the bound is optimal. The new lower bound also implies improved lower bounds for (k,z)(k,z)-clustering in doubling metrics. (2) For the upper bound, we provide efficient coreset construction algorithms for (k,z)(k,z)-clustering with improved or optimal coreset sizes in several metric spaces. In particular, we provide an O~z(k2z+2z+2ε2)\tilde{O}_z(k^{\frac{2z+2}{z+2}} \varepsilon^{-2})-sized coreset, with a unfied analysis, for (k,z)(k,z)-clustering for all z1z\geq 1 in Euclidean space. [1] Cohen-Addad, Larsen, Saulpic, Schwiegelshohn. STOC'22. [2] Cohen-Addad, Larsen, Saulpic, Schwiegelshohn, Sheikh-Omar, NeurIPS'22.

Keywords

Cite

@article{arxiv.2211.11923,
  title  = {On Optimal Coreset Construction for Euclidean $(k,z)$-Clustering},
  author = {Lingxiao Huang and Jian Li and Xuan Wu},
  journal= {arXiv preprint arXiv:2211.11923},
  year   = {2024}
}

Comments

Accepted by STOC '24

R2 v1 2026-06-28T06:25:40.599Z