Experimental Comparison of Greedy Randomized Adaptive Search Procedures for the Maximum Diversity Problem

Abstract. The maximum diversity problem (MDP) consists of identifying optimally diverse subsets of elements from some larger collection. The selection of elements is based on the diversity of their characteristics, calculated by a function applied on their attributes. This problem belongs to the class of NP-hard problems. This paper presents new GRASP heuristics for this problem, using different construction and local search procedures. Computational experiments and performance comparisons between GRASP heuristics from literature and the proposed heuristics are provided and the results are analyzed. The tests show that the new GRASP heuristics are quite robust and find good solutions to this problem.

• Type: Conference
• Year: 2004
• Where: WEA
• Authors: Geiza C. Silva, Luiz Satoru Ochi, Simone L. Martins
• Keywords: Algorithms, Diverse Subsets, GRASP Heuristics, Maximum Diversity Problem, WEA 2004