English

Random Projections for k-Means: Maintaining Coresets Beyond Merge & Reduce

Data Structures and Algorithms 2020-02-19 v5

Abstract

We give a new construction for a small space summary satisfying the coreset guarantee of a data set with respect to the kk-means objective function. The number of points required in an offline construction is in O~(kϵ2min(d,kϵ2))\tilde{O}(k \epsilon^{-2}\min(d,k\epsilon^{-2})) which is minimal among all available constructions. Aside from two constructions with exponential dependence on the dimension, all known coresets are maintained in data streams via the merge and reduce framework, which incurs are large space dependency on logn\log n. Instead, our construction crucially relies on Johnson-Lindenstrauss type embeddings which combined with results from online algorithms give us a new technique for efficiently maintaining coresets in data streams without relying on merge and reduce. The final number of points stored by our algorithm in a data stream is in O~(k2ϵ2log2nmin(d,kϵ2))\tilde{O}(k^2 \epsilon^{-2} \log^2 n \min(d,k\epsilon^{-2})).

Keywords

Cite

@article{arxiv.1504.01584,
  title  = {Random Projections for k-Means: Maintaining Coresets Beyond Merge & Reduce},
  author = {Marc Bury and Chris Schwiegelshohn},
  journal= {arXiv preprint arXiv:1504.01584},
  year   = {2020}
}

Comments

This paper has been withdrawn due to an error in Theorem 1

R2 v1 2026-06-22T09:11:36.643Z