Banca de DEFESA: CARLOS ANTONIO RAMÍREZ BELTRAN

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
STUDENT : CARLOS ANTONIO RAMÍREZ BELTRAN
DATE: 24/03/2021
TIME: 09:00
LOCAL: Remota - por vídeo conferência
TITLE:

Portability of Models by using Transfer Learning to predict undergraduate students performance

KEY WORDS:

Transfer Learning; Machine Learning; Student Performance; Moodle.

PAGES: 83
BIG AREA: Ciências Exatas e da Terra
AREA: Ciência da Computação
SUBÁREA: Metodologia e Técnicas da Computação
SPECIALTY: Sistemas de Informação
SUMMARY:

In recent years, one of the major challenges in education has been to predict students’ performance in a certain and reliable way in order to apply different strategies toimprove their academic deficiencies. In this way, there are many works focused on finding Machine Learning (ML) models individually, but few use the knowledge acquired from student data of a course degree to predict the students’ outcome of other curses’s degree. Therefore, the main goal of this work is to analise the portability of ML models by using Transfer Learning (TL) in students’ logs extracted form the Moodle system, aiming to predict undergraduate students performance in different courses. Through the applied experimental methodology, each of the two types of groupings formed by the disciplines
will be evaluated: those formed according to the undergraduate course and those according to the activities used in Moodle. The data extraction of each group will be performed from the Moodle logs, using the following evaluation methods: cross-validation and hold-out. With this, it will be possible to know whether these evaluations, all performed on the predictive models with the J48 algorithm, tend to show different results in relation to the portability of forecasting models. For evaluation, two scenarios were developed for the execution of experiments, so that each experiment consists of two parts: the choice of models, using the AUC ROC index for Experiment 1 and the F-Measure for Experiment 2; and the validation of the models, using the Precision index for Experiment 1 and the Recall for Experiment 2. The results, even in the evaluation phase, allow us to affirm that it is possible to apply transfer learning between models of the same group in some cases.

BANKING MEMBERS:
Presidente - 4351681 - JOAO CARLOS XAVIER JUNIOR
Interno - 1669545 - DANIEL SABINO AMORIM DE ARAUJO
Externa ao Programa - 2451906 - ADJA FERREIRA DE ANDRADE
Externo à Instituição - MARCELO DAMASCENO DE MELO - IFRN