|
The predictive model of higher education guidance using integrated techniques for imbalanced data of learner groups |
|---|---|
| รหัสดีโอไอ | |
| Creator | Atsawin Surawatchayotin |
| Title | The predictive model of higher education guidance using integrated techniques for imbalanced data of learner groups |
| Contributor | Worapat Paireekreng |
| Publisher | Faculty of Engineering and Technology, Mahanakorn University of Technology |
| Publication Year | 2564 |
| Journal Title | Journal of Information Science and Technology |
| Journal Vol. | 11 |
| Journal No. | 1 |
| Page no. | 65-74 |
| Keyword | Imbalanced data, Recursive Feature Elimination, SMOTE, Ensemble learning-based model,, Multiple-intelligence skills, Single Imputation |
| URL Website | https://tci-thaijo.org/index.php/JIST |
| Website title | Journal of Information Science and Technology |
| ISSN | 2651-1053 |
| Abstract | An inappropriate selection of higher education entrance came from insufficient experience of student and in-depth knowledge of the subject area. Most of students rely on social norms and values to make a decision without consideration of individual skill. Therefore, in order to select the appropriate subject area, the integrated prediction model of higher education entrance based on student's multiple-intelligence and skills is necessary. According to the research, there is the imbalance data of subject areas. Hence, the integrated techniques which combined Single Imputation, SMOTE, and feature Selection were used to create prediction model. It also implemented Bagging, Boosting and Stacking in the experiments to address the problems. The experimental results found that the most importance skill using the stacking technique appeared to subject area of mathematical logic. There is an accuracyrate of 77% performed better which is compared to other techniques. This subject is the fundamental of other knowledge areas. It affects the admission to higher education. |
