Comparison of Random Forest Algorithm, Support Vector Machine and Neural Network for Classification of Student Satisfaction Towards Higher Education Services

Abstract

Students are customers of a higher education institution. Various types of services and facilities are provided by universities for students, such as academic services, student services, and infrastructure including information technology services. Student satisfaction with the quality of the university services is one of the factors that affect the academic performance of the university. This study aims to compare the level of accuracy between the Random Forest method, Support Vector Machine and Neural Network for the classification of student satisfaction with the services provided by Higher Education. The dataset consisted of 430 respondents from private universities in Vietnam taken and adapted from previous research. The training and testing dataset uses a split proportion of 80:20 of the total available datasets. Model performance measurements using matrix confusion include accuracy, sensitivity (recall), and precision. The results showed that the Random Forest method can produce the best accuracy rate of 76.47%, followed by Neural Network and Support Vector Machine algorithms with an accuracy rate of 74.12% on each.