Optimization of Stress Classification Among Students Using Random Forest Algorithm

Abstract

Stress is a condition of physical and psychological discomfort experienced by students due to academic pressure, demands from parents and teachers, and schoolwork. This stress can lead to physical tension, behavioral changes, and mental health problems if not handled properly.  Random Forest is a promising approach to analyze and classify student stress. The aim of this study is to classify stress among students to enable the development of targeted interventions to support student well-being and academic success. The dataset used was sourced from Kaggle and included 1100 datasets with 21 columns. The research stages included data preprocessing, exploratory data analysis, modeling, Decision tree classification and evaluation of the confusion matrix model and Deployment as a measure of stress level. Classification results were evaluated by calculating accuracy, precision, recall and f1-score for stress classes (low, medium and high). The results of this study resulted in an accuracy value before tuning of 87.27% and after tuning of 88.64%. This research can provide insights for schools, parents, and government to develop more effective strategies in addressing the problem of stress among students. The use of Random Forest algorithm is proven to be effective in analyzing and classifying stress, so that it can help in decision making and appropriate welfare interventions to tackle before stress reaches critical levels.