Implementação de uma metodologia para mineração de dados aplicada ao estudo de núcleos convectivos / Implementation of methodology for data mining applied to the study of convective nucleous

AUTOR(ES)
DATA DE PUBLICAÇÃO

2005

RESUMO

In this work, a methodology for data mining was implemented using the rough sets theory and applied to the study of convective nucleous. Data mining has been used to analyze large volumes of data trying to identify frequent correlations, patterns, and outliers, in the most varied domains of applications, including scientific research. In this study, data mining was applied to a spatial-temporal database composed of occurrence data of electric discharge of the type cloud-to-ground, stability indexes obtained from radiosounding stations, and initialization data of the a mesoscale meteorological model. In face of the large amount of electric discharge data, a method for reducing these data was required. Several methods of spatial representation of data were investigated, in order to spatially group discharge occurrence data in entities that were named centers of electrical activity. This reduction allowed to identify patterns in a feasible amount of time, and made possible the integration with the remaining data. The objective of this work was to detect unknown and potentially useful information in the considered database and to demonstrate the potential of the proposed methodology. The results prove the feasibility of the developed tool.

ASSUNTO(S)

conjunto aproximativos convective system estimador de núcleo atmospheric electrical discharge mineração de dados meteorologia kernel estimator descargas elétricas atmosféricas meteorology rought data mining sistemas convectivos

ACESSO AO ARTIGO

Documentos Relacionados