English

Shortest Cycles With Monotone Submodular Costs

Data Structures and Algorithms 2022-11-10 v1 Computational Complexity

Abstract

We introduce the following submodular generalization of the Shortest Cycle problem. For a nonnegative monotone submodular cost function ff defined on the edges (or the vertices) of an undirected graph GG, we seek for a cycle CC in GG of minimum cost OPT=f(C)\textsf{OPT}=f(C). We give an algorithm that given an nn-vertex graph GG, parameter ε>0\varepsilon > 0, and the function ff represented by an oracle, in time nO(log1/ε)n^{\mathcal{O}(\log 1/\varepsilon)} finds a cycle CC in GG with f(C)(1+ε)OPTf(C)\leq (1+\varepsilon)\cdot \textsf{OPT}. This is in sharp contrast with the non-approximability of the closely related Monotone Submodular Shortest (s,t)(s,t)-Path problem, which requires exponentially many queries to the oracle for finding an n2/3εn^{2/3-\varepsilon}-approximation [Goel et al., FOCS 2009]. We complement our algorithm with a matching lower bound. We show that for every ε>0\varepsilon > 0, obtaining a (1+ε)(1+\varepsilon)-approximation requires at least nΩ(log1/ε)n^{\Omega(\log 1/ \varepsilon)} queries to the oracle. When the function ff is integer-valued, our algorithm yields that a cycle of cost OPT\textsf{OPT} can be found in time nO(logOPT)n^{\mathcal{O}(\log \textsf{OPT})}. In particular, for OPT=nO(1)\textsf{OPT}=n^{\mathcal{O}(1)} this gives a quasipolynomial-time algorithm computing a cycle of minimum submodular cost. Interestingly, while a quasipolynomial-time algorithm often serves as a good indication that a polynomial time complexity could be achieved, we show a lower bound that nO(logn)n^{\mathcal{O}(\log n)} queries are required even when OPT=O(n)\textsf{OPT} = \mathcal{O}(n).

Keywords

Cite

@article{arxiv.2211.04797,
  title  = {Shortest Cycles With Monotone Submodular Costs},
  author = {Fedor V. Fomin and Petr A. Golovach and Tuukka Korhonen and Daniel Lokshtanov and Giannos Stamoulis},
  journal= {arXiv preprint arXiv:2211.04797},
  year   = {2022}
}

Comments

17 pages, 1 figure. Accepted to SODA 2023

R2 v1 2026-06-28T05:29:48.149Z