A Fast Algorithm For The Greedy Interchange For Large-Scale Clustering And Median Location Problems

A fast algorithm is described for implementing the greedy interchange heuristic for lise in solving large scale clustering and uncapacitated median location problems. Computational experience is reported for these algorithms on a number of large randomly generated networks and on some difficult problem sets, and comparisons with some other implementations are made. An additional heuristic is proposed for solving these set partitioning problems based on an efficient procedure for achieving the interchange.