Genetic Algorithm with Fast Greedy Heuristic for Clustering and Location Problems

Authors propose new genetic algorithm for solving the planar p-median location problem and k-means clustering problem. The ideas of the algorithm are based on the genetic algorithm with greedy heuristic for the p-median problem on networks and information bottleneck (IB) clustering algorithms. The proposed algorithm uses the standard k-means procedure or any other similar algorithm for local search. The efficiency of the proposed algorithm in comparison with known algorithms was proved by experiments on large-scale location and clustering problems.