A polynomial matrix processing heuristic algorithm for finding high quality feasible solutions for the TSP

Gary Waissi, Pragya Kaushal

Research output: Contribution to journalArticle

Abstract

In this paper we present a simple heuristic algorithm to find high-quality, feasible solutions, for the traveling salesman problem (TSP). We hypothesize, that the quality of the initial solution provided by the proposed heuristic will improve the performance of the subsequent algorithm in terms of number of iterations required to reach a certain level TSP solution. The proposed heuristic does not attempt to compete against known TSP algorithms and heuristics, but instead, should be considered to serve as a “pre-processor”. The method provides a simple framework for testing new node selection and neighborhood rules. The cost matrix of origin and destination pairs is processed in a systematic way starting from a principal diagonal matrix element to find a feasible TSP tour. The matrix reduction, systematic moves in rows and columns, systematic elimination of rows and columns from further consideration, and the “reserved” column declaration, assure that the resulting sequence of nodes and edges forms a complete TSP tour. The process can be repeated from each principal diagonal element. The best TSP tour found can then be used, for example, as an input to another algorithm (e.g. the TABU search, simulated annealing, ant colony optimization, nearest neighbor, or another heuristic) to attempt to improve the tour further. It should be noted, that the proposed technique can also be used for testing of presence of cycles of a proposed solution provided by another algorithm. While the goal of the heuristic algorithm is to attempt to find the optimum tour, optimality cannot be guaranteed.

Original languageEnglish (US)
JournalOPSEARCH
DOIs
StateAccepted/In press - Jan 1 2019

Fingerprint

Traveling salesman problem
Heuristic algorithms
Polynomials
Processing
Ant colony optimization
Testing
Simulated annealing
Heuristic algorithm
Heuristics
Costs

Keywords

  • Traveling salesman problem
  • TSP
  • TSP algorithm
  • TSP heuristic

ASJC Scopus subject areas

  • Management Information Systems
  • Information Systems
  • Computer Science Applications
  • Management Science and Operations Research

Cite this

A polynomial matrix processing heuristic algorithm for finding high quality feasible solutions for the TSP. / Waissi, Gary; Kaushal, Pragya.

In: OPSEARCH, 01.01.2019.

Research output: Contribution to journalArticle

@article{d985927a130541428bb9ece7f98b338a,
title = "A polynomial matrix processing heuristic algorithm for finding high quality feasible solutions for the TSP",
abstract = "In this paper we present a simple heuristic algorithm to find high-quality, feasible solutions, for the traveling salesman problem (TSP). We hypothesize, that the quality of the initial solution provided by the proposed heuristic will improve the performance of the subsequent algorithm in terms of number of iterations required to reach a certain level TSP solution. The proposed heuristic does not attempt to compete against known TSP algorithms and heuristics, but instead, should be considered to serve as a “pre-processor”. The method provides a simple framework for testing new node selection and neighborhood rules. The cost matrix of origin and destination pairs is processed in a systematic way starting from a principal diagonal matrix element to find a feasible TSP tour. The matrix reduction, systematic moves in rows and columns, systematic elimination of rows and columns from further consideration, and the “reserved” column declaration, assure that the resulting sequence of nodes and edges forms a complete TSP tour. The process can be repeated from each principal diagonal element. The best TSP tour found can then be used, for example, as an input to another algorithm (e.g. the TABU search, simulated annealing, ant colony optimization, nearest neighbor, or another heuristic) to attempt to improve the tour further. It should be noted, that the proposed technique can also be used for testing of presence of cycles of a proposed solution provided by another algorithm. While the goal of the heuristic algorithm is to attempt to find the optimum tour, optimality cannot be guaranteed.",
keywords = "Traveling salesman problem, TSP, TSP algorithm, TSP heuristic",
author = "Gary Waissi and Pragya Kaushal",
year = "2019",
month = "1",
day = "1",
doi = "10.1007/s12597-019-00396-x",
language = "English (US)",
journal = "OPSEARCH",
issn = "0030-3887",
publisher = "Springer India",

}

TY - JOUR

T1 - A polynomial matrix processing heuristic algorithm for finding high quality feasible solutions for the TSP

AU - Waissi, Gary

AU - Kaushal, Pragya

PY - 2019/1/1

Y1 - 2019/1/1

N2 - In this paper we present a simple heuristic algorithm to find high-quality, feasible solutions, for the traveling salesman problem (TSP). We hypothesize, that the quality of the initial solution provided by the proposed heuristic will improve the performance of the subsequent algorithm in terms of number of iterations required to reach a certain level TSP solution. The proposed heuristic does not attempt to compete against known TSP algorithms and heuristics, but instead, should be considered to serve as a “pre-processor”. The method provides a simple framework for testing new node selection and neighborhood rules. The cost matrix of origin and destination pairs is processed in a systematic way starting from a principal diagonal matrix element to find a feasible TSP tour. The matrix reduction, systematic moves in rows and columns, systematic elimination of rows and columns from further consideration, and the “reserved” column declaration, assure that the resulting sequence of nodes and edges forms a complete TSP tour. The process can be repeated from each principal diagonal element. The best TSP tour found can then be used, for example, as an input to another algorithm (e.g. the TABU search, simulated annealing, ant colony optimization, nearest neighbor, or another heuristic) to attempt to improve the tour further. It should be noted, that the proposed technique can also be used for testing of presence of cycles of a proposed solution provided by another algorithm. While the goal of the heuristic algorithm is to attempt to find the optimum tour, optimality cannot be guaranteed.

AB - In this paper we present a simple heuristic algorithm to find high-quality, feasible solutions, for the traveling salesman problem (TSP). We hypothesize, that the quality of the initial solution provided by the proposed heuristic will improve the performance of the subsequent algorithm in terms of number of iterations required to reach a certain level TSP solution. The proposed heuristic does not attempt to compete against known TSP algorithms and heuristics, but instead, should be considered to serve as a “pre-processor”. The method provides a simple framework for testing new node selection and neighborhood rules. The cost matrix of origin and destination pairs is processed in a systematic way starting from a principal diagonal matrix element to find a feasible TSP tour. The matrix reduction, systematic moves in rows and columns, systematic elimination of rows and columns from further consideration, and the “reserved” column declaration, assure that the resulting sequence of nodes and edges forms a complete TSP tour. The process can be repeated from each principal diagonal element. The best TSP tour found can then be used, for example, as an input to another algorithm (e.g. the TABU search, simulated annealing, ant colony optimization, nearest neighbor, or another heuristic) to attempt to improve the tour further. It should be noted, that the proposed technique can also be used for testing of presence of cycles of a proposed solution provided by another algorithm. While the goal of the heuristic algorithm is to attempt to find the optimum tour, optimality cannot be guaranteed.

KW - Traveling salesman problem

KW - TSP

KW - TSP algorithm

KW - TSP heuristic

UR - http://www.scopus.com/inward/record.url?scp=85072036751&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85072036751&partnerID=8YFLogxK

U2 - 10.1007/s12597-019-00396-x

DO - 10.1007/s12597-019-00396-x

M3 - Article

AN - SCOPUS:85072036751

JO - OPSEARCH

JF - OPSEARCH

SN - 0030-3887

ER -