Decision Support System Using Classifiers for Predicting Student’s Academic Performance

The prediction based analysis over a dataset is one of the major data mining processing used by decision support system. In this present work such prediction based analysis is suggested to evaluate the student performance based on their academic background. The work is to classify the students based on the ranking to find out the weak student, average student and the top rank students. In this work comparative analysis will be performed by using different classifiers to perform the analysis on student data. The work will be implemented on management perspective to take necessary action for improving the performance of weak student and to improve the academic results. The work is suggested to be implemented in weka integrated java environment with user friendly features to select the appropriate classifier. Different classifiers are compared on weka on the basic of different parameters like time taken to build the model, correctly classified instances, accuracy, kappa statistic, mean absolute error, mean squared error, relative absolute error, relative squared error, Confusion matrix, TP rate & FP rate. Then the classifier for which the maximum number of instances are correctly and maximum accuracy is achieved is selected for developing decision support system.