Penerapan Seleksi Fitur Analysis of Variance Pada Algoritma Random Forest Classifier Dalam Klasifikasi Nilai Mahasiswa

Abstract

In educational institutions, such as universities, it is important to pay attention to the performance of students so that they can complete their studies on time. However, there are still issues where some students are unable to finish their studies on time, and some even decide to drop out or become inactive as students. This is evidenced by the decline in student grades in the Indonesian Language course from the 2020 to 2021 cohorts at UMKT. To address this problem, a method is needed to measure students' performance in completing their studies. This study aims to identify the attributes that influence the decline in student grades in the Indonesian Language course, as well as improve the accuracy of the Random Forest Classifier algorithm using ANOVA feature selection. The data used in this study consists of UMKT student data who took the Indonesian Language course from the 2020/2021 to 2021/2022 academic years. The data was obtained from the academic administration department (BAA) of UMKT and the General Basic Course Unit (MKDU) of UMKT, with a total of 1028 data points. The data analysis process was conducted using the 5-Fold Cross Validation method. The results of the study indicate that attributes such as Progress, % Course completed, Assignment 1, and Assignment 2 have a significant influence on the decline in student grades in the Indonesian Language course. Furthermore, the use of ANOVA feature selection in the Random Forest Classifier algorithm improved its performance, with the accuracy increasing from 85.65% to 87.31%.