|
The Study of Factors Affecting for On-time Graduation of Ungraduated Student Using Feature Selection Technique on Imbalanced Datasets |
|---|---|
| รหัสดีโอไอ | |
| Creator | Worakarn Jaidee |
| Title | The Study of Factors Affecting for On-time Graduation of Ungraduated Student Using Feature Selection Technique on Imbalanced Datasets |
| Contributor | Noppanat Wannapee |
| Publisher | Faculty of Information Science and Technology, Mahanakorn University of Technology |
| Publication Year | 2563 |
| Journal Title | Journal of Information Science and Technology |
| Journal Vol. | 10 |
| Journal No. | 1 |
| Page no. | 75-84 |
| Keyword | Imbalanced Datasets, Feature Selection, SMOTE |
| URL Website | https://tci-thaijo.org/index.php/JIST |
| Website title | Journal of Information Science and Technology |
| ISSN | 2651-1053 |
| Abstract | The objectives of this research were (1) to study the factors affecting thestudents' graduation according to the plan, and (2) to study the guidelines and improvements inteaching and learning problems of each subject affecting the graduation according to the plan ofthe student. Main factors or data in this research, the researcher used the grade point in eachsubject, GPA, the gender of the first-year and second-year students by collecting data fromComputer Information System, Faculty of Business Administration and Liberal Art,Rajamangala University of Technology Lanna, Chiang Mai with a total of 358 records. As aresult, it revealed that datasets were imbalanced with a higher number of students whograduated as their plan than students who did not graduate as their plan completely.Therefore, the researcher solved the problem using the method of balancing the data set bySynthetic Minority Over-Sampling Technique: SMOTE before entering the process ofanalyzing factors by using feature selection techniques which can divide as 3 techniques asfollows: (1) Chi-Square Feature Selection (2) Information Gain Feature Selection and (3)Correlation Based Feature Selection. From the result, it was found that the most important andtopnotch factor in every feature selection was Computer Programming 1 Subject This subject isan elementary and compulsory subject in the Computer Information System and most studentsreceived quite low points in this subject. Which may cause this subject to be topnotch in everyfeature selection techniques and maybe the main factor affecting graduation's plan of studentssignificantly. |
