Skip to main navigation Skip to search Skip to main content

On the Complexity of the Metric TSP under Stability Considerations

Research output: Chapter in Book/Report/Conference proceeding › Conference article in proceeding › Academic › peer-review

Abstract

We consider the metric Traveling Salesman Problem (Δ-TSP for short) and study how stability (as defined by Bilu and Linial [3]) influences the complexity of the problem. On an intuitive level, an instance of Δ-TSP is γ-stable (γ>1), if there is a unique optimum Hamiltonian tour and any perturbation of arbitrary edge weights by at most γ does not change the edge set of the optimal solution (i.e., there is a significant gap between the optimum tour and all other tours). We show that for γ ≥ 1.8 a simple greedy algorithm (resembling Prim’s algorithm for constructing a minimum spanning tree) computes the optimum Hamiltonian tour for every γ-stable instance of the Δ-TSP, whereas a simple local search algorithm can fail to find the optimum even if γ is arbitrary. We further show that there are γ-stable instances of Δ-TSP for every 1
Original languageEnglish
Title of host publicationSOFSEM 2011
Subtitle of host publicationTheory and Practice of Computer Science - 37th Conference on Current Trends in Theory and Practice of Computer Science, Proceedings
Pages382-393
Number of pages12
ISBN (Electronic)9783642183812
DOIs
Publication statusPublished - 2011
Externally publishedYes

Fingerprint

Dive into the research topics of 'On the Complexity of the Metric TSP under Stability Considerations'. Together they form a unique fingerprint.

Cite this