Neighborhood decomposition based variable neighborhood search and tabu search for maximally diverse grouping

The maximally diverse grouping problem (MDGP) is a relevant NP-hard optimization with number of real-world applications. However, solving large instances the computationally challenging. This work dedicated to new heuristic algorithm for problem, which distinguishes itself by two original features. First, it introduces first neighborhood decomposition strategy accelerate examinations. Second, integrates, in probabilistic way, complementary based local search procedures (variable descent and tabu search) as well an adaptive perturbation ensure suitable balance between intensification diversification space. Computational results on 320 benchmark commonly used literature show that proposed competes favorably state-of-the-art MDGP algorithms, reporting improved best-known (new lower bounds) 220 instances. Additional experiments are conducted analyze main components algorithm. can help better solve practical problems be formulated model.