English

Non-monotone submodular maximization under matroid and knapsack constraints

Computational Complexity 2009-02-03 v1 Data Structures and Algorithms

Abstract

Submodular function maximization is a central problem in combinatorial optimization, generalizing many important problems including Max Cut in directed/undirected graphs and in hypergraphs, certain constraint satisfaction problems, maximum entropy sampling, and maximum facility location problems. Unlike submodular minimization, submodular maximization is NP-hard. For the problem of maximizing a non-monotone submodular function, Feige, Mirrokni, and Vondr\'ak recently developed a 252\over 5-approximation algorithm \cite{FMV07}, however, their algorithms do not handle side constraints.} In this paper, we give the first constant-factor approximation algorithm for maximizing any non-negative submodular function subject to multiple matroid or knapsack constraints. We emphasize that our results are for {\em non-monotone} submodular functions. In particular, for any constant kk, we present a (1k+2+1k+ϵ)({1\over k+2+{1\over k}+\epsilon})-approximation for the submodular maximization problem under kk matroid constraints, and a (15ϵ)({1\over 5}-\epsilon)-approximation algorithm for this problem subject to kk knapsack constraints (ϵ>0\epsilon>0 is any constant). We improve the approximation guarantee of our algorithm to 1k+1+1k1+ϵ{1\over k+1+{1\over k-1}+\epsilon} for k2k\ge 2 partition matroid constraints. This idea also gives a (1k+ϵ)({1\over k+\epsilon})-approximation for maximizing a {\em monotone} submodular function subject to k2k\ge 2 partition matroids, which improves over the previously best known guarantee of 1k+1\frac{1}{k+1}.

Keywords

Cite

@article{arxiv.0902.0353,
  title  = {Non-monotone submodular maximization under matroid and knapsack constraints},
  author = {Jon Lee and Vahab Mirrokni and Viswanath Nagarjan and Maxim Sviridenko},
  journal= {arXiv preprint arXiv:0902.0353},
  year   = {2009}
}
R2 v1 2026-06-21T12:07:12.819Z