Improving Student Academic Performance Using an Attribute Selection Algorithm

Abstract

The aim of this paper is to predict the student academic performance using data mining techniques. Data mining is the process of extracting data and prediction. Prediction of student academic performance helps the students to take decision about their progress in academics. It can help not only the current students but also the future students to take decision. In this work, kaggle student dataset was used to predict the student performance. Different properties of the collected data were investigated and developed a classification hypothesis in order to apply data mining algorithms. A new attribute selection algorithm was proposed to identify the best attributes to increase the efficiency of student performance prediction. In this work, a machine learning tool called WEKA tool develop by the University of New Zealand was used for testing different algorithms on the data. The experimental results are validated against test data, and interesting co-relations are observed. The experimentation carried out with Naive Bayes, IBK, bagging, simple logistic, logistic, and random forest classifiers. The prediction results of student academic performance promising than most of the research work happened in educational data mining.