A Comparative Study of Classification Algorithms in EDM using 2 Level Classification for Predicting Student’s Performance

In higher education the performance of students is a most challenge work day by day in academic as well as in other curricular activities. As they all know that internet technology is growing as much as faster, but the learning approach of students are not up to the mark. The emerging research community which helps to find the solution to the said problem is Educational Data Mining. In present scenario, the huge students' data is stored in educational database. That type of database contains widely open or secret information to improve student performance. In our proposed work, we will have tested it on reputed dataset, which can be downloaded from a well known organization UCI repository and dataset name is student-mat.csv. This work has been investigated the process of classification of plethora of student’s data. Classification plot data into pre-determined groups of classes. It is often mentioned to as supervised learning because the classes are determined before analyzing the data. The work will to be divided into two parts. The first part will be the entropy based feature selection, after that classification process has to be performed. For the classification, we would have used 2 level classification method i.e, SVM and KNN. Later than observe the performance prediction of students based on parameters like accuracy, sensitivity, specificity of proposed method and is to be compared with some previous methods results.