Análise de dados sobre crianças com excesso de peso com auxílio de recursos computacionais

• Autor(a) principal: Bueno, Camilo Antonio Monteiro [UNESP]
• Data de Publicação: 2015
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UNESP
• Texto Completo: http://hdl.handle.net/11449/134144; http://www.athena.biblioteca.unesp.br/exlibris/bd/cathedra/19-01-2016/000856709.pdf
• topic: Nutrition; Nutrição; Obesidade nas crianças; Mineração de dados (Computação)

Childhood obesity has reached worrying rates in the Brazilian population; its occurrence has acquired great significance in the health field. Thus, in addition to the diagnosis of the obesity in children, other factors involved with its emergence should be investigated. The process of identification and analysis of these factors has proven to be a very complex task - the great number of studies published on this subject, many of them inconclusive, is an evidence of this complexity. This master thesis aimed to analyze data on children, especially overweight ones, in order to establish hypotheses on the role of some factors that may be involved with childhood obesity. Among the analyzed factors, some have been cited in the literature as possibly related to the emergence of the childhood obesity, namely, the delivery method, the infant weight at birth, the period of breastfeeding and the number of daily hours of sleep, in addition to the obesity hereditary. The factor pregnancy period was also included in the analysis. A second objective of this study was to test and evaluate the application of the association technique, from the data mining area, together with the database resource, as an auxiliary tool in the analysis process specially on the data types that are of interest in this work. In order to achieve the intended objectives, a set of data involving 107 children between 6 and 13 years old was analyzed. The data was collected in the city of Marília - SP and it consists of 38 normal weight children, 38 obese ones and 31 overweight ones. The data analysis process allowed to identify some relevant correlations between variables involved in the problem and establish hypotheses to be analyzed in more details. Regarding the used computational resources, association rules seem to be quite interesting for the analysis of the types of data considered in this study, since they allow to establish hypotheses from the correlations found between ...