The inapproximability of non-NP-hard optimization problems

  • Authors:
  • Liming Cai;David Juedes;Iyad Kanj

  • Affiliations:
  • School of Electrical Engineering and Computer Science, Ohio University, Athens, OH;School of Electrical Engineering and Computer Science, Ohio University, Athens, OH;School of Computer Science, Telecommunications, and Information Systems, DePaul University, Chicago, IL

  • Venue:
  • Theoretical Computer Science
  • Year:
  • 2002

Quantified Score

Hi-index 5.23

Visualization

Abstract

The inapproximability of non-NP-hard optimization problems is investigated. Techniques are given to show that the problems LOG DOMINATING SET and LOG HYPERGRAPH VERTEX COVER cannot be approximated to a constant ratio in polynomial time unless the corresponding NP-hard versions are also approximable in deterministic subexponential time. A direct connection is established between non-NP-hard problems and a PCP characterization of NP. Reductions from the PCP characterization show that LOG CLIQUE is not approximable in polynomial time and MAX SPARSE SAT does not have a FPTAS under the assumption that NP ⊈ DTIME(2O(√ n log n). A number of nontrivial approximation-preserving reductions are also presented, making it possible to extend inapproximability results to more natural non-NP-hard problems such as TOURNAMENT DOMINATING SET and RICH HYPERGRAPH VERTEX COVER.