Predicting Academic Performance Using Genetic Algorithm and SVM Classifier

Predicting the performance of a student is a great concern to the higher education managements, where several factors affect the performance. The scope of this paper is to investigate the accuracy of data mining techniques in such an environment. The first step of the study is to gather student’s data and their marks and technical, analytical, communicational and problem solving abilities. We collected records of 200 under graduate students of computer science course, from a private Educational Institution conducting various Under Graduate and Post Graduate courses. The second step is to extract and select the data and choose the relevant attributes for the accuracy using genetic algorithm. Attributes were classified into two groups “Demographic Attributes” and “Performance Attributes”. In the third step, support vector machine algorithms were constructed and their performances were evaluated. This work will help the institute to accurately predict the performance of the students for their growth.