Multicriteria global minimum cuts

Amitai Armon*, Uri Zwick

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

27 Scopus citations


We consider two multicriteria versions of the global minimum cut problem in undirected graphs. In the k-criteria setting, each edge of the input graph has k non-negative costs associated with it. These costs are measured in separate, non-interchangeable, units. In the AND-version of the problem, purchasing an edge requires the payment of all the k costs associated with it. In the OR-version, an edge can be purchased by paying any one of the k costs associated with it. Given k bounds b1,b2,. . . ,bk, the basic multicriteria decision problem is whether there exists a cut C of the graph that can be purchased using a budget of bi units of the ith criterion, for 1 ≤ i ≤ k. We show that the AND-version of the multicriteria global minimum cut problem is polynomial for any fixed number k of criteria. The OR-version of the problem, on the other hand, is NP-hard even for k = 2, but can be solved in pseudo-polynomial time for any fixed number k of criteria. It also admits an FPTAS. Further extensions, some applications, and multicriteria versions of two other optimization problems are also discussed.

Original languageEnglish
Pages (from-to)15-26
Number of pages12
Issue number1
StatePublished - Sep 2006


  • Graph algorithms
  • Minimum cut
  • Multicriteria optimization


Dive into the research topics of 'Multicriteria global minimum cuts'. Together they form a unique fingerprint.

Cite this