EBAFS-ETCs: Enhanced Bat Algorithm based Feature Selection and Ensemble Three Classifiers (ETCs) to Predict Student’s Academic Performance

C. John Paul and Dr.R. Santhi

Educational Data Mining (EDM) is a knowledge science, and a promising discipline, concerned with analyzing and studying data from academic databases. Through the investigation of these huge databases, using different data mining techniques, one can recognize distinctive models which assist research, classify and develop a student's educational performance. The main focus of higher learning institutions is to progress the overall quality and proficiency of teaching. Predicting the student’s behaviour and future in this world of information is the requirement of today in academic. In this work, Enhanced Bat Algorithm based Feature Selection and Ensemble Three Classifiers (EBAFS-ETCs) method is proposed to classify student’s performance depends on with selected features. Initially the samples are chosen from the knowledge repository which is pre-processed through Min Max Normalization (MMN) and Z Score Normalization (ZCN) method. Then the chosen attribute from EBAFS are linked to the learner’s communication along with the e-learning management scheme. The student’s analytical performance method is compared through group of classifiers, like Adaptive Neuro Fuzzy Inference System (ANFIS), Support Vector Machine (SVM) classifier and Decision Tree (DT). Furthermore, Ensemble Three Classifiers (ETCs) is proposed to develop the performance of these three classifiers. Bagging is the widespread ensemble approach which is utilized to merge the outcomes of the three classifiers. The obtained outcomes disclose that there is a strong relationship among student’s performance and their educational achievement.

Volume 11 | Issue 2

Pages: 268-280