Multiobjective fuzzy clustering approach based on tissue-like membrane systems

Fuzzy clustering problem is usually posed as an optimization problem. However, the existing researchhas shown that clustering technique that optimizes a single cluster validity index may not provide satisfactory results on different kinds of data sets. This paper proposes a multiobjective clustering...

ver descrição completa

Detalhes bibliográficos
Autores: Peng, Hong, Shi, Peng, Wang, Jun, Riscos Núñez, Agustín, Pérez Jiménez, Mario de Jesús
Formato: artículo
Estado:Versión enviada para evaluación y publicación
Fecha de publicación:2017
País:España
Recursos:Universidad de Sevilla (US)
Repositorio:idUS. Depósito de Investigación de la Universidad de Sevilla
OAI Identifier:oai:idus.us.es:11441/85116
Acesso em linha:https://hdl.handle.net/11441/85116
https://doi.org/10.1016/j.knosys.2017.03.024
Access Level:acceso abierto
Palavra-chave:Fuzzy clustering
Multiobjective clustering problem
Membrane Computing
Tissue-like membrane systems
Descrição
Resumo:Fuzzy clustering problem is usually posed as an optimization problem. However, the existing researchhas shown that clustering technique that optimizes a single cluster validity index may not provide satisfactory results on different kinds of data sets. This paper proposes a multiobjective clustering frameworkfor fuzzy clustering, in which a tissue-like membrane system with a special cell structure is designed tointegrate a non-dominated sorting technique and a modified differential evolution mechanism. Based onthe multiobjective clustering framework, a fuzzy clustering approach is realized to optimize three cluster validity indices that can capture different characteristics. The proposed approach is evaluated on sixartificial and ten real-life data sets and is compared with several multiobjective and singleobjective techniques. The comparison results demonstrate the effectiveness and advantage of the proposed approachon clustering the data sets with different characteristics.