Predicting Effects of Music in University Students Productivity Using Machine Learning

Abstract

Humans have used music as a fundamental aspect of their existence for centuries because it modifies their emotional state as well as their mental processing abilities and behavioral responses. Multiple investigations have analyzed the impact of musical listening on productivity especially within educational settings where students do their studies repeatedly. The effects of music on focus depend on different variables which include musical genres together with tempo and lyrics and personal emotional state. Students tend to enhance their concentration with instrumental and classical or jazz music but lyrics in songs can disrupt reading and writing academic tasks. Researchers conducted this investigation to explore the effects of music on university student performance and identity various music expressions and individual preferences on academic results. The researchers trained AdaBoost alongside Random Forest, Decision Tree, Logistic Regression, and SVC as machine learning models using a database which included data regarding student music preferences together with study habits along with productivity assessment results. A system of prediction models analyzed productivity levels by categorizing them into "High," "Medium," or "Low" along with the three input features of study hours, music type, and perceived productivity during musical sessions. The AdaBoostClassifier model reached the highest predictive accuracy of 98.51% but all algorithms produced results with relatively low accuracy than this model. The research indicates music impacts productivity levels but the effects show clear personal differences thus researchers need to improve both filter creation processes and system selection techniques for better accuracy. The discovery of complex music-to-productivity dynamics means future research needs to study multiple influencing variables more extensively.