Student Performance Classification using Support Vector Machine (SVM) with Polynomical Kernel on Online Student Activities

—The increasing usage of classification algorithms has encouraged researchers to explore many topics, including academic-related topics. In addition, the availability of data from various academic information management systems in recent years has been increasing, causing classification to become a technique that is in demand by educational institutes. Thereby, having a classification technique is important in researching the data on students’ performance. The purpose of this study is to classify students’ performance by using a polynomial kernel of Support Vector Machine (SVM) on online students’ activities. A new dataset is proposed in this study, which consists of academic and student online behaviours that influence the students’ performance. The proposed dataset also undergoes pre-processing stage to improve the accuracy and identify the significance of the proposed features. The experiment for SVM-POLY classification performance was set with a range of values on the parameters to be optimised by an optimisation algorithm, Grid Search. Classification accuracy, Precision, Recall and f1-score were applied to observe the result and determine the best classifier performance. The experimental results show that SVM – POLY, with a gamma value of 0.005, regularisation value of 0.1 and degree value of 1, come out with the best performance compared to a default value of SVM – POLY. The study is significant towards educational data mining in analysing the students’ performance during online students’ activities.