Learning Progress Modeling for Monitoring Student

Abstract

Monitoring the progress of students is part of the teacher's job which is very important and very time-consuming. Especially if there are many students with various subjects. This is the experience of most primary school teachers in Indonesia. One way to solve this problem is to predict student progress. In this study, the students' progress was predicted using Random Forest. The Random Forest algorithm is used because it can classify data that has incomplete attributes, which are usually found in student assessment data. The prediction model was built based on assessment data from 2 classes with 46 elementary school students in subjects: Indonesian, mathematics, SBdP (Cultural Arts and Crafts), PPKN (Pancasila and Citizenship Education), and Computers. The dataset comes from the formative and summative assessment results from 3 aspects (cognitive, psychomotor, and affective). The resulting model performance will be measured using accuracy and recall. The results showed that using a dataset of 5 subjects from 46 students, the Random Forest algorithm produced a learning progress model with 100% accuracy for training data and 94% for testing data. Meanwhile, the learning progress prediction model for each subject has 100% accuracy on training data and more than 96% on test data.