English

Approximating Low-Dimensional Coverage Problems

Computational Geometry 2011-12-06 v1 Data Structures and Algorithms

Abstract

We study the complexity of the maximum coverage problem, restricted to set systems of bounded VC-dimension. Our main result is a fixed-parameter tractable approximation scheme: an algorithm that outputs a (1\eps)(1-\eps)-approximation to the maximum-cardinality union of kk sets, in running time O(f(\eps,k,d)poly(n))O(f(\eps,k,d)\cdot poly(n)) where nn is the problem size, dd is the VC-dimension of the set system, and f(\eps,k,d)f(\eps,k,d) is exponential in (kd/\eps)c(kd/\eps)^c for some constant cc. We complement this positive result by showing that the function f(\eps,k,d)f(\eps,k,d) in the running-time bound cannot be replaced by a function depending only on (\eps,d)(\eps,d) or on (k,d)(k,d), under standard complexity assumptions. We also present an improved upper bound on the approximation ratio of the greedy algorithm in special cases of the problem, including when the sets have bounded cardinality and when they are two-dimensional halfspaces. Complementing these positive results, we show that when the sets are four-dimensional halfspaces neither the greedy algorithm nor local search is capable of improving the worst-case approximation ratio of 11/e1-1/e that the greedy algorithm achieves on arbitrary instances of maximum coverage.

Keywords

Cite

@article{arxiv.1112.0689,
  title  = {Approximating Low-Dimensional Coverage Problems},
  author = {Ashwinkumar Badanidiyuru and Robert Kleinberg and Hooyeon Lee},
  journal= {arXiv preprint arXiv:1112.0689},
  year   = {2011}
}
R2 v1 2026-06-21T19:45:47.212Z