Planar Graphs, Parallel Computing, and the Traveling Salesman Problem

Document Type

Presentation Abstract

Presentation Date

9-11-2003

Abstract

The traveling salesman problem, or TSP for short, asks for the minimum-weight Hamiltonian cycle in a edge-weighted complete graph. The most successful approach known for solving instances of the TSP is the linear-programming based method of Dantzig, Fulkerson, and Johnson from the early 1950s. In this talk we discuss some recent work on extending this approach to larger TSP instances. Following the work of Fleischer and Tardos (1999) and Letchford (2000), we consider computational techniques for exploiting graph planarity in improving the linear-programming relaxations. We present computational results with a parallel implementation of this approach and discuss some open theoretical issues that could have impact on the success of the method in practice.

Additional Details

This talk is part of The Big Sky Conference, and is sponsored in part by the National Science Foundation & the Department of Mathematical Sciences.

Thursday, 11 September 2003
4:10 p.m. in the UC Theatre

This document is currently not available here.

Share

COinS