Backpropagation Neural Network-Based Classification in Prioritizing Stress Attributes in Students Using the Analytic Hierarchy Process

Universities in Indonesia require final-year students to complete a thesis as a graduation requirement. However, many students face challenges in choosing a research topic, which can trigger stress and hinder timely graduation. The aim of this study is to identify the impact of stress caused by procrastination on students' timely graduation. In this research, the Analytic Hierarchy Process (AHP) method was used to determine the most influential procrastination attributes, while the Backpropagation Neural Network (BPNN) was employed to predict timely graduation. The results indicate that not working on the thesis is the most significant factor affecting student stress. Stress due to procrastination has a significant impact on graduation predictions, with an RMSE value of 0.1428 and an accuracy rate of 97.9%. These findings can assist educational institutions in designing more effective interventions to reduce stress caused by procrastination, thereby increasing the likelihood of timely graduation for students.