Uplift Modeling for preventing student dropout in higher education

Olaya, Diego; Vásquez, Jonathan; Maldonado, Sebastián; Miranda, J. Jaime; Verbeke, Wouter

doi:10.1016/j.dss.2020.113320

Cited by 68 publications

(51 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Dropout is a concern for higher education institutions because of the negative impact on the well-being of students and the community. The early withdrawal of students from degree programs has, on the one hand, monetary costs to educational institutions (i.e., loss of cash inflows) and, on the other hand, broad social costs [2]. Thus, understanding the causes of higher education dropout is essential to eliminate or at least minimise them as much as possible.…”

Section: Introductionmentioning

confidence: 99%

How Does Learning Analytics Contribute to Prevent Students’ Dropout in Higher Education: A Systematic Literature Review

Oliveira

Sobral

Ferreira

et al. 2021

BDCC

View full text Add to dashboard Cite

Retention and dropout of higher education students is a subject that must be analysed carefully. Learning analytics can be used to help prevent failure cases. The purpose of this paper is to analyse the scientific production in this area in higher education in journals indexed in Clarivate Analytics’ Web of Science and Elsevier’s Scopus. We use a bibliometric and systematic study to obtain deep knowledge of the referred scientific production. The information gathered allows us to perceive where, how, and in what ways learning analytics has been used in the latest years. By analysing studies performed all over the world, we identify what kinds of data and techniques are used to approach the subject. We propose a feature classification into several categories and subcategories, regarding student and external features. Student features can be seen as personal or academic data, while external factors include information about the university, environment, and support offered to the students. To approach the problems, authors successfully use data mining applied to the identified educational data. We also identify some other concerns, such as privacy issues, that need to be considered in the studies.

show abstract

Section: Introductionmentioning

confidence: 99%

How Does Learning Analytics Contribute to Prevent Students’ Dropout in Higher Education: A Systematic Literature Review

Oliveira

Sobral

Ferreira

et al. 2021

BDCC

View full text Add to dashboard Cite

show abstract

“…Por otra parte Himmel (2012), divide los enfoques del análisis de la deserción y retención en cinco grandes categorías, dependiendo del énfasis que otorgan a las variables explicativas, ya sea individuales, institucionales o del medio familiar, de la siguiente forma: psicológicos, económicos, sociológicos, organizacionales y de interacciones. Referente a los modelos de clasificación para la deserción en (Olaya et al, 2020), presentan un modelo Uplift en el cual se muestran los resultados y las virtudes del modelado de mejora en la adaptación de los esfuerzos de retención en la educación superior sobre los enfoques convencionales de modelado predictivo. Viloria et al, (2019), presentan un clasificador bayesiano aplicado a la deserción en la educación superior, la investigación propone un nuevo clasificador bayesiano simple (SBND) con Márkov de la variable de clase a una estructura de red donde se utiliza la herramienta "Weka" para realizar la clasificación y el modelo propuesto se compara estadísticamente con otros clasificadores bayesianos.…”

Section: Estado Del Arteunclassified

Análisis comparativo de técnicas de clasificación para determinar la deserción estudiantil de la facultad de ingeniería de la Universidad de Antioquia, Colombia

Chaparro¹,

Cuatindioy²,

Barrera³

2021

Espacios

View full text Add to dashboard Cite

El objetivo de este estudio es comparar diferentes algoritmos de clasificación de aprendizaje supervisado como las redes neuronales artificiales, métodos probabilísticos como regresión logística multinomial, métodos de ensamble como random forest, bagging, boosting y las máquinas de soporte vectorial, con el fin de identificar perfiles de posibles estudiantes desertores de la facultad de ingeniería de la Universidad de Antioquia, a partir de dos targets; número de créditos inscritos en último semestre y semestre en el cual el estudiante abandona la universidad.

show abstract

“…In [19], the students' attendance, also in the first semester, is analysed to predict dropout risk. In [20], the authors applies the uplift modeling framework to the problem of student dropout prevention and estimating the effects of actions such as tutorial with the objective to improve the design of retention programs. In [21], well-known data mining classification algorithms, predicts students' overall academic performance based on the information available at the end of the first year of the students' academic path at the University of Porto (Portugal).…”

Section: B Related Workmentioning

confidence: 99%

Creating a Recommender System to Support Higher Education Students in the Subject Enrollment Decision

et al. 2020

View full text Add to dashboard Cite

Higher Education plays a principal role in the changing and complex world of today, and there has been rapid growth in the scientific literature dedicated to predicting students' academic success or risk of dropout thanks to advances in Data Mining techniques. Degrees such as Computer Science in particular are in ever greater demand. Although the number of students has increased, the number graduating is still not enough to provide society with as many as it requires. This study contributes to reversing this situation by introducing an approach that not only predicts the dropout risk or students' performance but takes action to help both students and educational institutions. The focus is on maximizing graduation rates by constructing a Recommender System to assist students with their selection of subjects. In particular, the challenge is addressed of constructing reliable Recommender Systems on the basis of data which are both sparse and few in quantity, imbalanced, and anonymized, and which might have been stored under imperfect conditions. This approach is successfully applied to create a Recommender System using a realworld dataset from a public Spanish university containing performance data of a Computer Science degree course, demonstrating its successful application in real environments. The construction of a support system based on that approach is described, its results are evaluated, and its implications for students' academic achievement, and for institutions' graduation rates are discussed. Through the construction of this decision support system for students, we intend to increase the graduation rates and lower the dropout rate.

show abstract

Uplift Modeling for preventing student dropout in higher education

Cited by 68 publications

References 44 publications

How Does Learning Analytics Contribute to Prevent Students’ Dropout in Higher Education: A Systematic Literature Review

How Does Learning Analytics Contribute to Prevent Students’ Dropout in Higher Education: A Systematic Literature Review

Análisis comparativo de técnicas de clasificación para determinar la deserción estudiantil de la facultad de ingeniería de la Universidad de Antioquia, Colombia

Creating a Recommender System to Support Higher Education Students in the Subject Enrollment Decision

Contact Info

Product

Resources

About