High School Students’ Performance Analyzer Using Machine Learning Approach

A number of variables may contribute to student performance, students academic achievement predictions form a very important area of study in the field of educational research. The educational progress of students depends on a number of factors during his or her high school education, and they are not solely determined by exam results class discussion or assignments. This work offers a new approach to estimating student grades by using both machine learning and PLS-SEM to improve the prediction efficiency. First, self-administered questionnaire was administered to selected high school teachers. In order to establish which factors they consider have the greatest impact on student performance. By applying PLS-SEM, we highlighted some of the most influential attributes: pre-term exam scores, term exams, assignments, class attendance, group projects, and classroom participation. These factors were chosen. Out from the wide purview of factors it makes the scope narrower for the analysis. Based on the key factors, it was followed by the prediction of student grades using machine learning techniques including Random Forest, Decision Tree, and Support Vector Regression. This approach gives a more accurate and reliable prediction of student outcomes since the machine Learning is involved together with the multivariate nature of PLS. SEM that has the capacity to identify significant variables. The The findings of this investigation provide educators and policymakers with crucial new information on the significant determinants of achievement. Such a strategy could achieve more accurate identification of children who require intervention and might improve academic outcomes for all students in high school.