Proceedings of the Annual Meeting of the Cognitive Science Society

This article used an empirical experiment and a computationalmodel to test the hypothesis that humans rely on the visualsystem to solve the traveling salesperson problem (TSP). Wetested two consequences of this hypothesis: (1) humans shouldperform better on Euclidean TSP than not–Euclidean TSP; (2)a model of the visual system should account for performance inEuclidean TSP. Participants were asked to solve Euclidean ornot–Euclidean TSP, and a pyramid model of the visual systemwas used to solve the same tours as the humans. The resultsshow that deviations from the optimal tour were smaller in Eu-clidean problems than in not–Euclidean problems, and the fitof the pyramid model to human performance was worse onnot–Euclidean problems then on Euclidean problems. Theseresults suggest that participants solve Euclidean problems withthe visual system, but that other mechanisms are needed to suc-cesfully solve non–visual problems.