A Real-Life Machine Learning Experience for Predicting University Dropout at Different Stages Using Academic Data

Abstract

High levels of school dropout are a major burden on the educational and professional development of a country's inhabitants. A country's prosperity depends, among other factors, on its ability to produce higher education graduates capable of moving a country forward. To alleviate the dropout problem, more and more institutions are turning to the possibilities that artificial intelligence can provide to predict dropout as early as possible. The difficulty of accessing personal data and privacy issues that it entails force the institutions to rely on the Academic Data of their students to create accurate and reliable predictive systems. This work focuses on creating the best possible predictive model based solely on academic data, and accordingly, its capacity to infer knowledge must be maximised. Thus, Feature Engineering and Instance Engineering techniques such as dealing with redundancy, significance of the features, correlation, cardinality features, missing values, creation or elimination of features, data fusion, removal of unuseful instances, binning, resampling, normalisation, or encoding are applied in detail before the construction of well-known models such as Gradient Boosting, Random Forest, and Support Vector Machine along with an Ensemble of them at different stages: prior to enrolment, at the end of the first semester, at the end of the second semester, at the end of the third semester, and at the end of the fourth semester. Through the construction of these predictive models that serve as inputs to a decision support system, the application of effective dropout prevention policies can be applied.

Effectiveness of a Peer Mentoring on University Dropout and Academic Performance

“…Different models have attempted to explain university dropout through theories such as attribution, expectations, goal setting, self-efficacy, or positive psychology (Aljohani, 2016;Demetriou & Schmitz-Seiborski, 2011;Flores et al, 2022). Additionally, new technological advances, such as machine learning, have been applied to the study of university dropouts (Cannistrà et al, 2021;Fernández-García et al, 2021;Opazo et al, 2021;Palacios et al, 2021).…”

Fulfilling K-12 Students' Mathematics Needs: Near Peer Teaching and Scientific Intelligent Clustering Tutor Approach

“…The effectiveness of curriculum characteristics and learning performance prediction models can be obtained by spiking feedforward neural algorithms [25]. The application of feature engineering and instance engineering techniques can detect over 72% of students at risk of dropping out [26]. Data mining processed by ML can be a new process and innovation to improve student and teacher learning performance, serving as a tool to collect, analyze, measure, and report student performance aimed at understanding and optimizing learning [27].…”

Forecasting Students Dropout: A UTAD University Study

“…Fernádez-García et al [10] defined several models from enrollment up to the fourth semester using mainly academic data. The approach considered the output of previous stages, i.e., each step assumed the prior knowledge generated.…”

“…Usually, the other data was not registered by UTAD's Staff. Therefore, the working dataset curricular units (courses) are identified with the acronym CU in Table 2 and its marks ranges [10,20] if the students succeeds or 0 if fails. Figure 1 plots the pairwise relationships between the most important curricular units obtained in Section 3.6.…”