Performance Evaluation of Various Classifiers in Analyzing Student Academic Progression under Factors Influencing Learning Attitudes

Abstract

Analyzing student academic performance plays an active role in educational organizations. The academic performance being anticipated underneath number of factors. Most of the researchers are concentrating on analyzing the performance primarily based on Cumulative Grade Point Average (CGPA), early academic records, student behavior etc. the learning attitudes of students are mostly influenced by various factors that are considered under three headings such as family or social environment, personal and educational institution environment. This paper analyzes a range of factors that impact the learning attitudes of students. Based on these factors, the academic performance of student will also be varied. This paper constructs four classifier models to analyze the overall performance of students primarily under the factors influencing their learning attitudes. Four predictor models had been used in the experimentation; they had been J48, SMO, multilayerPerceptron and Naïve Bayes classifiers. From the statistics obtained from the experimentation it is clear that multilayerPerceptron classifier model does the correct classification with a percentage of 94.7115 % which is highly accurate compared to remaining models.