Clustering under Perturbation Stability in Near-Linear Time
Abstract
We consider the problem of center-based clustering in low-dimensional Euclidean spaces under the perturbation stability assumption. An instance is -stable if the underlying optimal clustering continues to remain optimal even when all pairwise distances are arbitrarily perturbed by a factor of at most . Our main contribution is in presenting efficient exact algorithms for -stable clustering instances whose running times depend near-linearly on the size of the data set when . For -center and -means problems, our algorithms also achieve polynomial dependence on the number of clusters, , when for any constant in any fixed dimension. For -median, our algorithms have polynomial dependence on for in any fixed dimension; and for in two dimensions. Our algorithms are simple, and only require applying techniques such as local search or dynamic programming to a suitably modified metric space, combined with careful choice of data structures.
Cite
@article{arxiv.2009.14358,
title = {Clustering under Perturbation Stability in Near-Linear Time},
author = {Pankaj K. Agarwal and Hsien-Chih Chang and Kamesh Munagala and Erin Taylor and Emo Welzl},
journal= {arXiv preprint arXiv:2009.14358},
year = {2020}
}