PERFORMANCE EVALUATION OF 7TH GRADE STUDENTS FOR SOCIAL SCIENCE EDUCATION (IPS) UTILISING SUPPORT VECTOR MACHINE (SVM) METHOD

Abstract

In this study, a Support Vector Machine (SVM) method was utilized to predict the 7^th grade performance of social science education (IPS) within the following advanced levels and to delineated an evaluation of ongoing teaching plans for 7^th grade teachers. The model dataset was built for 192 students, consisting of cognitive and psychomotor formative The dataset refers to three classification categories (Adequate, Qualified, and Skilled) employed in computational algorithms for processing using linear and non-linear (polynomial and gaussian). The SVM model performance evaluation results obtained a performance accuracy (ACC) of 84% (linear), 75% for polynomials, and gaussians (90%), respectively. The Mathew Correlation Coefficient (MCC) evaluation described a validated performance of 47% for linear, 40% and 20% for polynomial and gaussians, respectively. In conclusion, student performances can follow the learning optimally at the next level, while teachers can replicate the learning process for 7th grade in future classrooms.