Open AccessComparative Analysis of the Learning on KDD Cup 2015 Dataset
Author(s) -
S. Nithya,
S. Umarani
Publication year - 2022
Publication title -
webology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.259
H-Index - 18
ISSN - 1735-188X
DOI - 10.14704/web/v19i1/web19050
Subject(s) - dropout (neural networks) , learning analytics , computer science , analytics , educational data mining , machine learning , artificial intelligence , process (computing) , set (abstract data type) , measure (data warehouse) , data mining , programming language , operating system
Massive open online courses (MOOCs) have delivered a high level of education around the world, but the significant dropout rate has impacted their educational efficiency. In MOOC, the researchers mainly focused on dropout prediction using various approaches. Our work emphasizes the real-time dataset of KDD CUP 2015 to extract certain main features, implemented by using various methods, and find out the different performance measures with different metrics. To measure early learner dropout, this work used a set of unique features extracted from real-time datasets. Individual user contributions and performance are frequently analyzed using the learning analytics framework. It also aids in the decision-making process for students in higher education. According to the data, the suggested model has a classification accuracy of 78 to 85%. The experimental results are also predicted, which one of the learning methods is mostly applicable to the dropout prediction in MOOC. In this evaluation, the performance measures are compared and the best one is reported.