2020 · 10 citations · 16 references
EducationComputer ScienceHigher EducationData MiningEducational Data MiningMachine LearningAdmissions AnalyticsDropout PredictionStudent Enrollment AnalysisStudent Risk IdentificationAccuracyAdmission ControlData-driven Admissions Decision SupportNaive BayesNaive Bayes ClassificationNearest-neighbor ClassificationStudent Risk ClassificationUniversity Admissions Systems
One of the most frequent problems in undergraduate level is pertaining to students' academic performance in terms of the length of study period and also the overall Grade Point Average (GPA). In many engineering-related departments, these indicators are still below the university standards. To overcome this problem, a tool for enrolling freshmen into university must be built using data gathered before the students are admitted. This research proposes a solution for these problems by means of Educational Data Mining to identify those who are at risk of failing during their study. Three different methods were investigated to classify the patterns, namely k-Nearest Neighbor, Artificial Neural Network, and Naïve Bayes. The datasets were collected by extracting the students' admission data. The experiment showed that the Artificial Neural Network demonstrated the best accuracy and precision. Therefore, it might become an appropriate Decision Support System for university admissions to enroll freshmen in the near future.
16
Evandro Costa, Baldoíno Fonseca, Marcelo Almeida Santana et al. · Computers in Human Behavior · 2017 · 411 citations
Early dropout prediction using data mining: a case study with high school students
Carlos Márquez‐Vera, Alberto Cano, Cristóbal Romero et al. · Expert Systems · 2015 · 311 citations