cs.DS

# Title: Approximation Algorithms for Continuous Clustering and Facility Location Problems

Abstract: We consider the approximability of center-based clustering problems where the points to be clustered lie in a metric space, and no candidate centers are specified. We call such problems "continuous", to distinguish from "discrete" clustering where candidate centers are specified. For many objectives, one can reduce the continuous case to the discrete case, and use an $\alpha$-approximation algorithm for the discrete case to get a $\beta\alpha$-approximation for the continuous case, where $\beta$ depends on the objective: e.g. for $k$-median, $\beta = 2$, and for $k$-means, $\beta = 4$. Our motivating question is whether this gap of $\beta$ is inherent, or are there better algorithms for continuous clustering than simply reducing to the discrete case? In a recent SODA 2021 paper, Cohen-Addad, Karthik, and Lee prove a factor-$2$ and a factor-$4$ hardness, respectively, for continuous $k$-median and $k$-means, even when the number of centers $k$ is a constant. The discrete case for a constant $k$ is exactly solvable in polytime, so the $\beta$ loss seems unavoidable in some regimes.
In this paper, we approach continuous clustering via the round-or-cut framework. For four continuous clustering problems, we outperform the reduction to the discrete case. Notably, for the problem $\lambda$-UFL, where $\beta = 2$ and the discrete case has a hardness of $1.27$, we obtain an approximation ratio of $2.32 < 2 \times 1.27$ for the continuous case. Also, for continuous $k$-means, where the best known approximation ratio for the discrete case is $9$, we obtain an approximation ratio of $32 < 4 \times 9$. The key challenge is that most algorithms for discrete clustering, including the state of the art, depend on linear programs that become infinite-sized in the continuous case. To overcome this, we design new linear programs for the continuous case which are amenable to the round-or-cut framework.
 Comments: 24 pages, 0 figures. Full version of ESA 2022 paper this https URL . This version adds a link to the conference version and fixes minor formatting issues Subjects: Data Structures and Algorithms (cs.DS) Journal reference: 30th Annual European Symposium on Algorithms (ESA 2022), 2022, pages 33:1--33:15 Cite as: arXiv:2206.15105 [cs.DS] (or arXiv:2206.15105v3 [cs.DS] for this version)

## Submission history

From: Ankita Sarkar [view email]
[v1] Thu, 30 Jun 2022 08:12:17 GMT (489kb)
[v2] Wed, 6 Jul 2022 08:55:14 GMT (37kb)
[v3] Thu, 1 Sep 2022 19:01:05 GMT (492kb)

Link back to: arXiv, form interface, contact.