Table 6.

Accuracy results for artificial neural networks (ANNs).

Study	Year	Predictive features	Accuracy
[17]	2015	Gender, location, type of school, high school score, CGPA, number of credits, and results	84.6%
[39]	2016	Test mark, class and lab performance, attendance, assignment, study time, previous result, family education, living area, drug addiction, affair, social media, and final year results	88%
[18]	2016	Online quizzes, email communication, content creation, and content interaction	98.3%
[22]	2018	Grades, gender, nationality, place of birth, section ID, topic, raised hand, discussion, class in 1st and 2nd terms, attendance, and parent satisfaction	85.4%,
[23]	2018	Gender, attendance, results, economic status, and parental education	-
[24]	2019	Gender, CGPA, English, Chinese, math, science, and proficiency test	84.8%
[25]	2019	Gender, content score, time spent, homework score, and attendance	80.5%
[46]	2019	CourseID, total of learning sessions, length of sessions, total of assessments of semester 1, grades, quizzes, and emails sent	97.4%
[26]	2019	Gender, nationality, place of birth, StageID, GradeID, SectionID, topic, semester, relation, raised hands, discussion, parent survey and satisfaction, and attendance	73.5%
[36]	2020	Gender, age, address location, parent job, travel time, study time, free time, failures, activities, health, and abstance	64.40%
[56]	2021	Gender, region, educational level, age range, neighborhood crime rate (IMD), number of times they have previously participated in the course, enrolled credits, disability, and the final exam result (passed/failed). In addition, the number of times the student has interacted with any of the online course contents has been counted throughout the courses	78.20%
[63]	2020	Gender, content score, time spent, number of entries to content, homework score, attendance, and archived courses	80.47%
[57]	2021	123 variables	82.10% (high) 70.89% (low)
[58]	2021	116 features for the production and 84 for the learning phase	80.76% and 86.57%