An Empirical Analysis of Approximation Algorithms for Euclidean TSP

- The Traveling Salesman Problem (TSP) is perhaps the most famous optimization problem in the set NP-hard. Many problems that are natural applications in computer science and engineering can be modeled using TSP and therefore, researchers are searching implementations focusing on the quality of the produced solution and their computational time. An innovative Polynomial-Time Approximation Scheme (PTAS) for the Euclidean TSP was given by Sanjeev Arora. To date, there is no evidence that it can be implemented to be practically useful. In light of this, we propose an implementation of the Euclidean TSP that is based on the essential steps of the Arora's PTAS and adds some heuristics to improve the running time. The paper includes the description of a C++ Arora-based implementation and a comparative analysis of this implementation with various competitive algorithms for the Euclidean TSP.

• Type: Conference
• Year: 2009
• Where: CSC
• Authors: Bárbara Rodeker, M. Virginia Cifuentes, Liliana Favre
• Keywords: CSC 2007, CSC 2009, Euclidean TSP, Famous Optimization Problem, Traveling Salesman Problem