CHSMST+: An algorithm for spatial clustering

• Autor(a) principal: Valencio, Carlos Roberto [UNESP]
• Data de Publicação: 2017
• Outros Autores: Medeiros, Camila Alves [UNESP], Neves, Leandro Alves [UNESP], Zafalon, Geraldo Francisco Donega [UNESP], De Souza, Rogeria Cristiane Gratao [UNESP], Colombini, Angelo Cesar
• Tipo de documento: Artigo de conferência
• Idioma: eng
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://dx.doi.org/10.1109/PDCAT.2016.081; http://hdl.handle.net/11449/178996
• topic: CHSMST (Clustering based on Hyper Surface and Minimum Spanning Tree); Hyper Surface Classification (HSC); Minimum Spanning Tree (MST); Spatial Clustering; Spatial Data Mining

Spatial clustering has been widely studied due to its application in several areas. However, the algorithms of such technique still need to overcome several challenges to achieve satisfactory results on a timely basis. This work presents an algorithm for spatial clustering based on CHSMST, which allows: data clustering considering both distance and similarity, enabling to correlate spatial and nonspatial data, user interaction is not necessary, and use of multithreading technique to improve the performance. The algorithm was tested ia a real database of health area.