Estimasi Keberhasilan Siswa dalam Pemodelan Data Berbasis Learning Menggunakan Algoritma Support Vector Machine

Abstract

SMK Negeri 5 Pekanbaru aims to prepare competent graduates who can compete in the global market. The realization of these goals is influenced by student achievement at school. Student achievements determine the ability of students to work in certain fields. Based on observations, it is known that student achievement at SMK Negeri 5 Pekanbaru tend to be low. This is also shown by the data that has been collected through the Curriculum section. Based on the data, there can be extraction using the supervised learning method to make a classification model of student achievements. The supervised learning algorithm used in this research is a Support Vector Machine (SVM). The data used in this study are student's data grade X SMK Negeri 5 Pekanbaru in 2020 totaling 160 data. The classification process is carried out by applying the GridSearch method to find the best kernel to be implemented. Based on the implementation of GridSearch, the kernel to be used is Radial Basis Function (RBF) with Cost (C) and Gamma (?) parameters. Based on 16 experiments with different parameter values, the best classification results are obtained using the value of Cost (C) = 0.1 and the value of Gamma (?) = 0.01, with accuracy values of 94%.