A hierarchical approach for solving large-scale traveling salesman problem

A hierarchical approach which combines an unsupervised learning algorithm and a recursive Hopfield neural network is proposed for solving the large-scale traveling salesman problem (TSP). For a TSP with given number of cities, an unsupervised learning algorithm was first used to find the clusters for decomposing the given problem and then a recursive version of the Hopfield network was applied to the centroids of the clusters. The proposed recursive Hopfield network was also applied to cities in each cluster in order to find an optimal path. A final tour was obtained by connecting together the resultant paths of each cluster.<
>