An Analysis of the Fitness Landscape of Travelling Salesman Problem

Abstract

The fitness landscape of the travelling salesman problem is investigated for 11 different types of the problem. The types differ in how the distances between cities are generated. Many different properties of the landscape are studied. The properties chosen are all potentially relevant to choosing an appropriate search algorithm. The analysis includes a scaling study of the time to reach a local optimum, the number of local optima, the expected probability of reaching a local optimum as a function of its fitness, the expected fitness found by local search and the best fitness, the probability of reaching a global optimum, the distance between the local optima and the global optimum, the expected fitness as a function of the distance from an optimum, their basins of attraction and a principal component analysis of the local optima. The principal component analysis shows the correlation of the local optima in the component space. We show how the properties of the principal components of the local optima change from one problem type to another.

Inferring Future Landscapes: Sampling the Local Optima Level

“…The QAP is NP-Hard and manifests in many realworld problems; it remains a testbed for new fitness landscape analysis (Tayarani-N and Prügel-Bennett, 2016;Verel et al, 2018;Ochoa and Herrmann, 2018). This reality combined with the availability of QAP LON construction algorithms lead us to select QAP for this study.…”

Applications of artificial intelligence in battling against covid-19: A literature review

“…There has been some works that use evolutionary algorithms to solve them. Studying the fitness landscape of these problems can be helpful to better understand these problems and thus develop more successful algorithms [649] , [650] , [651] , [652] . Thus one area of research for future works can be studying the fitness landscape of these problems.…”

What an “Ehm” Leaks About You: Mapping Fillers into Personality Traits with Quantum Evolutionary Feature Selection Algorithms

“…In such a condition, further exploration of the solutions' space can be effective -meaning that the uncertainty above can be removed -only if another local minimum can be reached. For this reason, the core-idea of the PCA-QEA, is to ensure that the search through the solutions' space tends to explore to a finer scale the directions that the PCA shows to carry most of the variance [68], [69].…”