Home // PATTERNS 2015, The Seventh International Conferences on Pervasive Patterns and Applications // View article

A Rule for Combination of Spatial Clustering Methods

Authors:
Danielly Holmes
Ronei Moraes
Rodrigo Vianna

Keywords: Majority vote rule; Spatial clustering methods; Statistical significance.

Abstract:
In the area of spatial analysis, spatial clustering methods use georeferencing information in order to identify significant and non-significant spatial clusters of the phenomenon in study in a specific geographical region. Several methods are available in the literature, such as scan statistic, Getis-Ord statistics, and the Besag and Newell method. In practical applications, all those methods are not able to produce results which can capture the real event with good accuracy. In this paper, we propose using the a combining classifier technique in order to provide better results for spatial clustering methods, using the majority voting rule for that combination. A study case was presented using epidemiological data of dengue fever from state of Paraíba, Brazil, in the year of 2011 and the final results allowed to identify the priority and non-priority areas in the region of interest.

Pages: 55 to 59

Copyright: Copyright (c) IARIA, 2015

Publication date: March 22, 2015

Published in: conference

ISSN: 2308-3557

ISBN: 978-1-61208-393-3

Location: Nice, France

Dates: from March 22, 2015 to March 27, 2015