A Novel Data Mining Approach to Improve Learning Capacity and Academic Success in High-Risk Postgraduate Students in Distance Learning

Abstract

Introduction: Data mining (DM) is valuable in numerous academic disciplines. Since student data can be used to recognize important concept associated to student education behavior, education study is increasing rapidly. Educational Data mining (EDM) can assist institution to distinguish students’ successes by analyze through achievement.
Methods: This study address the classification of a DM methods, the Dynamic Particle Swarm Optimized- Discrete Support Vector Machine (DPSO-DSVM), to advance the knowledge ability and academic attainment of high-risk post-Graduate (PG) students in distance learning program. The Gujarat University PG student dataset was gathered. Clean, transform, reduce and Feature select the data.
Results: The research used the DPSO-DSVM algorithm to improve the model's prediction performance and its capacity to detect at-risk students in their academic careers. Performance indicators such as precision, recall, loss and accuracy were examined to determine the DPSO-DSVM method’s superiority in predicting academic performance over conventional techniques.
Conclusions: This innovative DM method can improve the educational results of high-risk PG distance learning students.