Revision #2 Authors: Piotr Berman, Marek Karpinski

Accepted on: 1st June 2006 00:00

Downloads: 3562

Keywords:

We design a polynomial time 8/7-approximation algorithm for the Traveling Salesman Problem in which all distances are either one or two. This improves over the best known approximation factor for that problem. As a direct application we get a 7/6-approximation algorithm for the Maximum Path Cover Problem,similarly improving upon the best known approximation factor for that problem. The result depends on a new method of consecutive path cover improvements and on a new analysis of certain related color alternating paths. This method could be of independent interest.

Revision #1 Authors: Piotr Berman, Marek Karpinski

Accepted on: 16th November 2005 00:00

Downloads: 1747

Keywords:

We design a polynomial time 8/7-approximation algorithm for the Traveling Salesman Problem in which all distances are either one or two. This improves over the best known approximation factor for that problem. As a direct application we get a 7/6-approximation algorithm for the Maximum Path Cover Problem,similarly improving upon the best known approximation factor for that problem. The result depends on a new method of consecutive path cover improvements and on a new analysis of certain related color alternating paths. This method could be of independent interest.

TR05-069 Authors: Piotr Berman, Marek Karpinski

Publication: 11th July 2005 16:35

Downloads: 1925

Keywords:

We design a polynomial time 8/7-approximation algorithm for the Traveling Salesman Problem in which all distances are either one or two. This improves over the best known approximation factor of 7/6 for that problem. As a direct application we get a 7/6-approximation algorithm for the Maximum Path Cover Problem, similarily improving upon the best known approximation factor for that problem. The result depends on a new method of consecutive path cover improvements and on a new analysis of certain related color alternating paths. This method could be of independent

interest.