Detecção de clusters irregulares para dados pontuais através danão-conectividade ponderada de grafos

Detalhes bibliográficos
Ano de defesa: 2011
Autor(a) principal: Angelica Ferreira Carvalho
Orientador(a): Não Informado pela instituição
Banca de defesa: Não Informado pela instituição
Tipo de documento: Dissertação
Tipo de acesso: Acesso aberto
Idioma: por
Instituição de defesa: Universidade Federal de Minas Gerais
UFMG
Programa de Pós-Graduação: Não Informado pela instituição
Departamento: Não Informado pela instituição
País: Não Informado pela instituição
Palavras-chave em Português:
Link de acesso: http://hdl.handle.net/1843/BUOS-8GHK7L
Resumo: Strategies for detecting clusters for both spatial regions data and for point data are already quite widespread, it is understood by data point, situati- ons in which each element in the population is treated individually, knowing its location on the map under study. The problems with irregularly shaped clusters are not closed. The most likely cluster generally spreads in large por- tions of the map, impacting its geographic significance. Statistical methods that use the Kulldorffs Spatial Scan, combined with penalty functions were used to control the excessive freedom of clusters shapes. These methods have been not applied to point data. In this context, we will present a novel multi- objective algorithm using the Spatial Scan Statistic and penalty function forNon-connectivity Weighted to points data. The solution is a Pareto set, con- sisting of all clusters not less in both objectives than the others. The best solution is determined by evaluating the significance through Monte-Carlo simulations. We use a statistical theory to evaluate the statistical significance of the solutions obtained by multi-objective algorithm that employs the concept of attainment functions.