The Machine Learning-Based Dropout Early Warning System for Improving the Performance of Dropout Prediction

Lee, Sunbok; Chung, Jae Young

doi:10.3390/app9153093

Cited by 111 publications

(67 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Through a series of tests, the authors identified that the key attributes that affect student dropouts are mostly academic aspects, namely performance grade, quarterly exam rating, attendance percentage, and written score (e.g., quizzes, summative test). Such results are similar to the past data of Pereira and Zambrano [4], wherein the primary causes of student dropout attributed to the academic variables such as low academic grade. Also, the location or distance of the school from the students' residences has related factors of students' dropout.…”

Section: Resultssupporting

confidence: 90%

“…Sunbok Lee and Jae Young Chung [4] introduced a study aimed at improving the efficiency of the early warning system: (a) dealing with the problem of class imbalance using synthetic minority oversampling techniques (SMOTE) and set-up approaches in machine learning; and (b) testing qualified classifiers with both receiver operating characteristics (ROC) and precision-recall (PR) curves. Towards this purpose, they trained random forest, improved decision tree, random forest with SMOTE, and upgraded decision tree with SMOTE using huge data samples from 165,715 high school students from the National Education Information System (NEIS) in South Korea.…”

Section: Related Literaturementioning

confidence: 99%

See 1 more Smart Citation

Predicting Students' Dropout Indicators in Public School using Data Mining Approaches

Gil¹

2020

IJATCSE

View full text Add to dashboard Cite

Currently, one of the challenges of academic institutions is dropout student issues. It is critical to recognize which students are at risk of losing in school and what are the underlying factors of dropout. The study of students' dropout in Lawigan National High School enables teachers to identify the influential factors of dropout cases in school preemptively, to respond to it immediately, and to assist prospective dropout students in continuing studies and gaining knowledge for a better future. Lawigan National High School is a public school supervised with the Department of Education. The said school holds all the records of the students. The study aims to identify the underlying factors of dropout students that need to have an intervention to lessen the value of dropouts. The Weka Experiment Environment platform used to run simulations with machine learning algorithms from ready datasets. Two classification algorithms C4.5 and Naïve Bayes (NB) tested on a dataset containing student academic and demographic details, e.g., age, gender, marital status, written, performance, and quarterly examination grade, attendance percentage, home distance, parents' income, mothers' education, fathers' education, and dropout status, using a 10-fold cross-validation to estimate generalization accuracy. The experiment result shows that the top indicator for students' dropout cases is the academic performance with 98.9474% accuracy as perceived by the C4.5 algorithm, as shown in Table 2.

show abstract

Section: Resultssupporting

confidence: 90%

Section: Related Literaturementioning

confidence: 99%

Predicting Students' Dropout Indicators in Public School using Data Mining Approaches

Gil¹

2020

IJATCSE

View full text Add to dashboard Cite

show abstract

“…As described in different systematic reviews [4,5], many models can be applied to education. Students' performance [6], students' dropout within an individual course [7,8], program retention [9], recommender systems in terms of activities [10], learning resources, [11] and next courses to be enrolled [12,13] are some examples of the application of those models. Independently of the desired outcome, models have used many different types of data in order to perform the predictions.…”

Section: Predictive Modelsmentioning

confidence: 99%

An Early Warning System to Detect At-Risk Students in Online Higher Education

2020

View full text Add to dashboard Cite

Artificial intelligence has impacted education in recent years. Datafication of education has allowed developing automated methods to detect patterns in extensive collections of educational data to estimate unknown information and behavior about the students. This research has focused on finding accurate predictive models to identify at-risk students. This challenge may reduce the students’ risk of failure or disengage by decreasing the time lag between identification and the real at-risk state. The contribution of this paper is threefold. First, an in-depth analysis of a predictive model to detect at-risk students is performed. This model has been tested using data available in an institutional data mart where curated data from six semesters are available, and a method to obtain the best classifier and training set is proposed. Second, a method to determine a threshold for evaluating the quality of the predictive model is established. Third, an early warning system has been developed and tested in a real educational setting being accurate and useful for its purpose to detect at-risk students in online higher education. The stakeholders (i.e., students and teachers) can analyze the information through different dashboards, and teachers can also send early feedback as an intervention mechanism to mitigate at-risk situations. The system has been evaluated on two undergraduate courses where results shown a high accuracy to correctly detect at-risk students.

show abstract

“…These can allow faculties to improve the learning/teaching strategies and interactions to be employed in virtual courses. It may also allow researchers to make predictions about the students' behavior during their process of learning, such as preventing dropouts in courses [47].…”

mentioning

confidence: 99%

Students’ Acceptance and Tracking of a New Container-Based Virtual Laboratory

et al. 2020

View full text Add to dashboard Cite

Presently, the ever-increasing use of new technologies helps people to acquire additional skills for developing an applied critical thinking in many contexts of our society. When it comes to education, and more particularly in any Engineering subject, practical learning scenarios are key to achieve a set of competencies and applied skills. In our particular case, the cybersecurity topic with a distance education methodology is considered and a new remote virtual laboratory based on containers will be presented and evaluated in this work. The laboratory is based on the Linux Docker virtualization technology, which allows us to create consistent realistic scenarios with lower configuration requirements for the students. The laboratory is comparatively evaluated with our previous environment, LoT@UNED, from both the points of view of the students’ acceptance with a set of UTAUT models, and their behavior regarding evaluation items, time distribution, and content resources. All data was obtained from students’ surveys and platform registers. The main conclusion of this work is that the proposed laboratory obtains a very high acceptance from the students, in terms of several different indicators (perceived usefulness, estimated effort, social influence, attitude, ease of access, and intention of use). Neither the use of the virtual platform nor the distance methodology employed affect the intention to use the technology proposed in this work.

show abstract

The Machine Learning-Based Dropout Early Warning System for Improving the Performance of Dropout Prediction

Cited by 111 publications

References 26 publications

Predicting Students' Dropout Indicators in Public School using Data Mining Approaches

Predicting Students' Dropout Indicators in Public School using Data Mining Approaches

An Early Warning System to Detect At-Risk Students in Online Higher Education

Students’ Acceptance and Tracking of a New Container-Based Virtual Laboratory

Contact Info

Product

Resources

About