With the increase in Massive Open Online Courses (MOOC) and Virtual Learning Environments (VLE), a general intuition about the student performance degradation is always attributed to the incorporation of technology in academic system. Procrastination is one of the perennial challenges faced by students that affects the overall academic performance. Computer-based learning environments have enhanced the overall academic system, but it sometimes a challenge for students to stick to assigned tasks and avoid procrastination. This is a comprehensive study conducted to uncover students' behavior based on their interactions with the VLE. Information of students from versatile educational arenas including science, technology, engineering, mathematics, and social sciences have been included to gain understanding of student behavior to classify student behavior as probable to procrastinate or not. The study uses different unsupervised machine learning techniques including k-means, mini-batch k-means, agglomerative and hierarchical clustering to identify the correlations. Astonishing facts revealed by the study, representing little to absolutely no correlation between the final scores and the activities performed while interacting with the computer-based learning environments. The study also emphasizes on the need of detailed logs of students' activity and properly annotated data to get valued insights for procrastination and academic performance of an individual, which would be helpful to identify students at risk of adverse academic performances and would lead to assist such students to improve their performance before off-track. © 2022 American Institute of Physics Inc.. All rights reserved.