<![CDATA[This study aims to classify attenuation quality in fiber optic networks using the K-Means algorithm. The classification model is designed to group data based on the characteristics of network parameters collected from field measurements. This research also compares the performance of models built with all features and those using selected features. The evaluation uses two main metrics, Sum of Squared Errors (SSE) and Silhouette Coefficient, to assess the quality of the clustering results. Testing shows that the use of selected features produces better clusters, with an SSE of 91.75 and a Silhouette Coefficient of 0.5923, compared to using all features, which results in an SSE of 101.98 and a Silhouette Coefficient of 0.4977. These results indicate that selecting the right features can positively impact cluster quality, although it is not the sole determining factor. Overall, K-Means has been proven effective in grouping attenuation quality, making it useful for supporting efficient fiber optic network maintenance and monitoring.]]>
Copyrights © 2025
