SEGMENTASI PENDENGAR MUSIK BERDASARKAN KEBIASAAN DAN DAMPAKNYA TERHADAP KESEHATAN MENTAL MENGGUNAKAN K-MEANS DAN RANDOM FOREST

Abstract

Music plays a crucial role in human life, not only as a form of entertainment but also as a tool for managing emotions and mental health. This study aims to analyze the relationship between music listening habits and individuals' psychological conditions, as well as to identify the key factors influencing the effects of music on mental health. The methods used in this research include K-Means Clustering to group music listeners based on their habits and mental states, and Random Forest to determine the dominant factors affecting their music experience. The findings reveal that music listeners can be categorized into four main clusters, with anxiety and insomnia emerging as the most dominant factors influencing music effects. The Silhouette Score of 0.24 indicates a moderately effective clustering performance, though there is room for improvement. The evaluation of the Random Forest model shows the highest accuracy at 86% in Cluster 3, while Cluster 0 records the lowest accuracy at 55%, suggesting variations in pattern recognition across clusters. This study concludes that music impacts each group of listeners differently, depending on their psychological characteristics. These findings can serve as the foundation for developing personalized music recommendation systems based on mental health conditions, as well as advancing music therapy for individuals with high levels of anxiety or sleep disorders.