Título: Um Novo Método para Predição da Evasão Escolar Usando Redes Neurais
Autores: Martinho, Valquíria R. C.; Nunes, Clodoaldo; Minussi, Carlos R.
Resumo: Dropping out of school is one of the most complex and crucial problems in education. Worldwide, the phenomenon afflicts and concerns everyone involved, from basic education to higher education, both public and private, causing social, economic, political, academic and financial losses. Therefore, it is fundamental to develop efficient methods for prediction, assessment and monitoring of the students at risk of dropping out, making the planning and adoption of proactive actions possible to minimize the situation. In this perspective, this paper presents the potentials of an intelligent, bold and innovative system, developed for the prediction of risk groups of student dropout in higher education classroom using a Fuzzy ARTMAP Neural Network, one of the techniques of artificial intelligence, with the possibility of continued learning. It was implemented in the technology courses of the Federal Institute of Mato Grosso, based on the academic and socioeconomic records of the students. The results, showing a success rate of the dropout group around 97% and overall accuracy over 76%, highlights the reliability and accuracy of the system. Furthermore, it is noteworthy that the strength and boldness of this research lies in the possibility of identifying early the eminent school dropout using only the enrollment data.
Palavras-chave: Higher education; school dropout; prediction dropout; artificial neural networks (ANN); Art Family; Fuzzy ARTMAP neural network
Código DOI: 10.21528/CBIC2013-003
Artigo em pdf: bricsccicbic2013_submission_3.pdf
Arquivo BibTex: bricsccicbic2013_submission_3.bib