Work

Educational Data Mining using ML

Year: 2023
Type: article
Abstract: : The ability to forecast students' performance is one of the most useful and important academic issues in the world today because of the development of technology. In the subject of education, data mining is incredibly useful, particularly for analysing student performance. The imbalanced datasets in this subject have made it extremely difficult to estimate students' performance, and there is no comparison of the various resampling techniques. This study compares multiple resampling procedures to manage the unbalanced information problem when projecting student performance of two distinctive datasets, including Borderline SMOTE, SMOTE-ENN, SMOTE, Random Over Sampler, SVM-SMOTE, and SMOTE-Tomek. Additionally, the dissimilarity between binary classification and multiclass, as well as the features' structures, are looked at. must be able to evaluate the effectiveness.
Language: English
Open Access status: diamond
OpenAlex ID: W4383342878
FWCI:
Cites: 7
Cited by:
Related to: 10