Comparative analysis of solving traveling salesman problem using artificial intelligence algorithms

• Main Authors: Brucal, Stanley Glenn E., Dadios, Elmer P.
• Format: text
• Published: Animo Repository 2017
• Subjects: Artificial intelligence; Algorithms; Computer software—Development; Traveling salesman problem; Electrical and Computer Engineering; Electrical and Electronics
• Online Access: https://animorepository.dlsu.edu.ph/faculty_research/1562; https://animorepository.dlsu.edu.ph/context/faculty_research/article/2561/type/native/viewcontent
• Institution: De La Salle University

This paper aims to provide a comparative study of the different artificial intelligence (AI) algorithms applied to solve the traveling salesman problem (TSP). Four (4) AI algorithms such as genetic algorithm, nearest neighbor, ant colony optimization, and neuro-fuzzy are executed in MatLab software to determine which among these techniques will provide the least execution time to solve a TSP. The objective of comparing and analyzing each AI algorithm - as applied to a single problem with the different program execution - is to identify if significant difference in execution time could lead to significant saving in energy consumption. The simulations using MatLab resulted to strong correlation at an R2 of 0.95 in the average execution time with the number of code lines, but do not give a significant execution time variance as when ANOVA and t-test measures were performed. The result of this paper could be used as a basis in the design phase of software development life cycle to arrive into an energy efficient software application with respect to time needed to execute a program. © 2017 IEEE.