Predict Student's Academic Performance and Evaluate the Impact of Different Attributes on the Performance Using Data Mining Techniques

In Educational Data Mining, many data mining techniques are applied to extract hidden knowledge from student data. This knowledge helps educational institutes to improve their teaching quality. Different attributes have an impact on student's academic performance. Behavioral and student absent in class has an effect on their academic performance. In this paper, the effect of above two categories of features will be measured using some data mining techniques. For classification Naïve Bayes, Artificial Neural Network, Decision Tree, and K-Nearest Neighbor algorithms are used in this paper. We also use ensemble methods like Bagging, Adaboosting, and Random Forest to get more accuracy. An ensemble filtering technique has been applied to detect misclassified instance from training dataset which improves the accuracy of predicting. In this paper, ensemble filtering technique gives the best accuracy of 84.3 percent where adaboosting on Artificial Neural Network gives 78.6 percent accuracy.