<![CDATA[Thread is one of the main raw materials in the production process of textile companies. The availability of thread consumption data in textile companies can be used to determine the pattern of thread consumption in a certain period. Data mining clustering method is one technique that can be used to form patterns from the data thread. In this study, hybrid Fuzzy C-Means (FCM) and Fuzzy Particle Swarm Optimization (FPSO) clustering algorithms are used, which are the combination of algorithms FCM and FPSO. This hybrid algorithm is able to overcome the weaknesses of the original algorithm, namely FCM. The purpose of this study is to test the performance of the FCM-FPSO hybrid method by implementing the clustering of the thread data from PT. Batam Bersatu Apparel (PT. BBA) into an application. The application implemented Unified Process for software engineering method. In this application, the performance of three methods was compared, those methods are FCM, FPSO and Hybrid FCM-FPSO. The result of the implementation is the lowest average objective function is 3441.00 achieved by the Hybrid FCM-FPSO algorithm, then followed by the FCM algorithm with value of 3540.33 and the highest is achieved by the FPSO algorithm with value of 4485.40. This result showed that the application was successfully proved that the FCM-FPSO Hybrid algorithm can produce the best thread clusters compared to the original method, FCM and FPSO.]]>
Copyrights © 2020
