A Comparative Study of Three Evolutionary Algorithms Incorporating Different Amounts of Domain Knowledge for Node Covering Problem

Research output: Contribution to journalArticle

Colleges, School and Institutes

Abstract

This paper compares three different evolutionary algorithms for solving the node covering problem: EA-I relies on the definition of the problem only without using any domain knowledge, while EA-II and EA-III employ extra heuristic knowledge. In theory, it is proven that all three algorithms can find an optimal solution in finite generations and find a feasible solution efficiently; but none of them can find the optimal solution efficiently for all instances of the problem. Through experiments, it is observed that all three algorithms can find a feasible solution efficiently, and the algorithms with extra heuristic knowledge can find better approximation solutions, but none of them can find the optimal solution to the first instance efficiently. This paper shows that heuristic knowledge is helpful for evolutionary algorithms to find good approximation solutions, but it contributes little to search for the optimal solution in some instances.

Details

Original languageEnglish
Pages (from-to)266-271
Number of pages6
JournalIEEE Transactions on Systems, Man and Cybernetics, Part C
Volume35
Issue number2
Publication statusPublished - 1 May 2005

Keywords

  • performance analysis, optimization methods, algorithm design, heuristic knowledge