Please use this identifier to cite or link to this item: http://repositorio.ufla.br/jspui/handle/1/12354
Title: Detecção de agrupamentos espaço-temporais de ocorrências de dengue utilizando processos pontuais
Other Titles: Detection of spatio-temporal clusters of dengue occurrences using point processes
Authors: Scalon, João Domingos
Lima, Renato Ribeiro de
Nogueira, Denismar Alves
Keywords: Dengue – Diagnóstico – Métodos estatísticos
Dengue – Três Corações, MG
Análise por agrupamento
Dengue – Diagnosis – Statistical methods
Dengue – Três Corações, State of Minas Gerais, Brazil
Cluster analysis
Issue Date: 21-Feb-2017
Publisher: Universidade Federal de Lavras
Citation: ABREU, R. F. de. Detecção de agrupamentos espaço-temporais de ocorrências de dengue utilizando processos pontuais. 2017. 73 p. Dissertação (Mestrado em Estatística e Experimentação Agropecuária)-Universidade Federal de Lavras, Lavras, 2017.
Abstract: Dengue is one of the most infectious diseases affecting the world's population, where around 50 million people get the disease every year and, approximately, 2.5 billion people are in risky areas. Brazil is one of the countries where the population is most prone to be infected with dengue. Understanding the spatial and temporal behavior of dengue cases is one of the most important aspects for the decision making of public health managers. Thus, the aim of this work was to present and evaluate several statistical methods that can be used to detect the presence of space-time clusters in dengue cases. The following methods are presented for the detection of global spatiotemporal clustering: Knox test, Mantel test, Jacquez test, homogeneous K function and non-homogeneous K function. The Scan statistic was also used to detect clusters at specific times and locations. The performance of the methods was evaluated from the application of them in data of occurrences of dengue in the city of Três Corações - MG, during the period from 01/01/2010 to 12/31/2015. The Knox, Mantel and Jacques tests indicated the presence of spatio-temporal clusters in dengue occurrences in the study region. From the analyzes using the homogeneous and non-homogeneous K functions, it was possible to verify that the patterns of clustering of dengue occurrences are results of first order effects (intensity) and not of second order effects (spatio-temporal dependence). The analysis with scan statistic allowed the identification of six significant local spatio-temporal clusters in the city of Três Corações. The results show that each method has its peculiarities and, therefore, should not be used individually for the detection of space-time clusters of dengue cases. It is recommended to use the combined methods for a more precise description of the spatio-temporal clustering of dengue cases.
URI: http://repositorio.ufla.br/jspui/handle/1/12354
Appears in Collections:Estatística e Experimentação Agropecuária - Mestrado (Dissertações)



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.