| Autor: |
Gustavo Silva Semaan, Augusto Cesar Fadel, Flávio Montenegro, José André de Moura Brito |
| Rok vydání: |
2021 |
| Předmět: |
|
| Zdroj: |
IEEE Latin America Transactions. 19:652-659 |
| ISSN: |
1548-0992 |
| DOI: |
10.1109/tla.2021.9448548 |
| Popis: |
This paper introduces two heuristic algorithms for the Maximum-Diameter Clustering Problem (MDCP), based on the Biased Random-Key Genetic Algorithm (BRKGA) and the Greedy Randomized Adaptive Search Procedure (GRASP) metaheuristics. This problem consists of finding k clusters that minimize the largest within-cluster distance (diameter) among all clusters. The MDCP is classified as NP-hard and presents increased difficulty when attempting to determine the optimal solution for any instance. The results obtained in the experiments using 50 well-known instances indicate a good performance of proposed heuristics, that outperformed both three algorithms and an integer programming model from the literature. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
