Classification of Student Academic Performance Based on Randomized and Synthetic Dataset

Analyzing the students’ performance has been a challenging task in the past and many data mining tools have come to derive the knowledge hidden in these data and WEKA is one such tool. In this work, a student performance standard benchmark dataset from UCI machine learning repository is analyzed using standard data mining classifiers like Bayes Net, J48, ID3,PART,LMT and REP Tree. The accuracy obtained from original data is compared with synthetic and Randomized dataset generated and this work proves that the accuracy of synthetic and Randomized data increases. This study will support educational decision makers to design the courses more effectively.