Approximation algorithms for multi-criteria traveling salesman problems

From MaRDI portal
Publication:1017906

DOI10.1007/11970125_24zbMATH Open1161.90019arXivcs/0606040OpenAlexW1902458947MaRDI QIDQ1017906FDOQ1017906

Bodo Manthey, L. Shankar Ram

Publication date: 13 May 2009

Published in: Algorithmica, Approximation and Online Algorithms (Search for Journal in Brave)

Abstract: In multi-criteria optimization problems, several objective functions have to be optimized. Since the different objective functions are usually in conflict with each other, one cannot consider only one particular solution as the optimal solution. Instead, the aim is to compute a so-called Pareto curve of solutions. Since Pareto curves cannot be computed efficiently in general, we have to be content with approximations to them. We design a deterministic polynomial-time algorithm for multi-criteria g-metric STSP that computes (min{1 +g, 2g^2/(2g^2 -2g +1)} + eps)-approximate Pareto curves for all 1/2<=g<=1. In particular, we obtain a (2+eps)-approximation for multi-criteria metric STSP. We also present two randomized approximation algorithms for multi-criteria g-metric STSP that achieve approximation ratios of (2g^3 +2g^2)/(3g^2 -2g +1) + eps and (1 +g)/(1 +3g -4g^2) + eps, respectively. Moreover, we present randomized approximation algorithms for multi-criteria g-metric ATSP (ratio 1/2 + g^3/(1 -3g^2) + eps) for g < 1/sqrt(3)), STSP with weights 1 and 2 (ratio 4/3) and ATSP with weights 1 and 2 (ratio 3/2). To do this, we design randomized approximation schemes for multi-criteria cycle cover and graph factor problems.


Full work available at URL: https://arxiv.org/abs/cs/0606040




Recommendations




Cites Work


Cited In (12)





This page was built for publication: Approximation algorithms for multi-criteria traveling salesman problems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1017906)