Banca de QUALIFICAÇÃO: MANOEL ISAC MAIA JUNIOR

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
DISCENTE : MANOEL ISAC MAIA JUNIOR
DATA : 30/04/2019
HORA: 10:30
LOCAL: Sala 222
TÍTULO:

EXPLORATORY DATA ANALYSIS OF UNIVERSITY RANKING USING NEURAL NETWORKS.

PALAVRAS-CHAVES:

Measurement of Performance. Data analysis. Self-Organizing Maps. Analysis of Clusters. Higher education institutions.

PÁGINAS: 45
GRANDE ÁREA: Engenharias
ÁREA: Engenharia de Produção
RESUMO:

The world lives the era of knowledge, new technologies and scientific advances, and as a result of a changing scenario, countries see in the university the possibility of being included in the world circuit of knowledge and skills. The evaluation of universities and higher education courses has been the focus of several rankings around the world, including Times High Education (THE) and the Ranking Universitário Folha (RUF). The rankings are an example of influence in the paradigm shift of the universities, since the positive results coming from the evaluation process give these institutions a reputation and social and academic prestige. In general, information on teaching, research and extension, internationalization activities and market evaluation are combined to generate a grade in which scales and weights are assigned for each dimension. The rankings serve a variety of purposes, including a benchmark for institutions about services and practices provided in addition to the strengths and weaknesses of each course / university. Self-organizing maps (SOM) are models of competitive neural networks. Through unsupervised learning, they perform a mapping between high-dimensional data for a display, usually two-dimensional, that brings the original density of information closer together, being a technique extensively used in areas such as data analysis and pattern recognition. With the use of post-processing techniques such as clustering of neurons, it is possible to generate regions associated with the original groupings of the data. This paper presents a spatiotemporal analysis of data from Brazilian universities through the training of maps with Ranking Universitário Folha data from 2014 and 2018. From the grouping profiles, after the segmentation of the trained maps, it will be possible to identify the positive and negative points of each grouping. With the identification of Higher Education Institutions (HEIs) in these different years, an analysis of the transitions between 2014 and 2018 will be carried out. The profiles of the groups are shown aiming to characterize the behavior of the groupings and the academic organizations, during the analyzed period. As a new alternative for analyzing the performance data of the HEI, the study also allows verifying disparities between the regions of Brazil and discusses some possible opportunities for improvement for the institutions.

MEMBROS DA BANCA:
Externo ao Programa - 346605 - GUTEMBERGUE SOARES DA SILVA
Interno - 1229030 - HELIO ROBERTO HEKIS
Presidente - 1142787 - JOSE ALFREDO FERREIRA COSTA