English

Linear-Time Approximation Scheme for k-Means Clustering of Affine Subspaces

Computational Geometry 2021-06-29 v1 Data Structures and Algorithms

Abstract

In this paper, we present a linear-time approximation scheme for kk-means clustering of \emph{incomplete} data points in dd-dimensional Euclidean space. An \emph{incomplete} data point with Δ>0\Delta>0 unspecified entries is represented as an axis-parallel affine subspaces of dimension Δ\Delta. The distance between two incomplete data points is defined as the Euclidean distance between two closest points in the axis-parallel affine subspaces corresponding to the data points. We present an algorithm for kk-means clustering of axis-parallel affine subspaces of dimension Δ\Delta that yields an (1+ϵ)(1+\epsilon)-approximate solution in O(nd)O(nd) time. The constants hidden behind O()O(\cdot) depend only on Δ,ϵ\Delta, \epsilon and kk. This improves the O(n2d)O(n^2 d)-time algorithm by Eiben et al.[SODA'21] by a factor of nn.

Keywords

Cite

@article{arxiv.2106.14176,
  title  = {Linear-Time Approximation Scheme for k-Means Clustering of Affine Subspaces},
  author = {Kyungjin Cho and Eunjin Oh},
  journal= {arXiv preprint arXiv:2106.14176},
  year   = {2021}
}
R2 v1 2026-06-24T03:38:12.730Z