<![CDATA[This study aims to compare the performance of three clustering methods k-Means++, DBSCAN (Density-Based Spatial Clustering of Applications with Noise), and Agglomerative Clustering in grouping data obtained from the Depression Anxiety Stress Scale (DASS-21) questionnaire. The questionnaire data represent three primary psychological dimensions: depression, anxiety, and stress. Performance evaluation was conducted by calculating the degree of agreement between the clustering results and the actual classes determined based on DASS-21 scores. The analysis results show that k-Means++ achieved the highest agreement rate of 35.63%, followed by Agglomerative Clustering at 34.39% and DBSCAN at 26.81%. These findings indicate that k-Means++ is more effective in forming clusters that align with the psychological dimensions present in the DASS-21 data. However, the relatively low agreement rates suggest overlapping emotional dimensions and the inherent complexity of psychological data. Therefore, further research is required through parameter optimization, data preprocessing, or the application of hybrid approaches to improve the quality and representativeness of the clustering results. ]]>
Copyrights © 2026
