Augmentation of the reconstruction performance of Fuzzy C-Means with an optimized fuzzification factor vector

Information granules have been considered as the fundamental constructs of Granular Computing. As a useful unsupervised learning technique, Fuzzy C-Means (FCM) is one most frequently used methods to construct information granules. The FCM-based granulation–degranulation mechanism plays pivotal role in In this paper, enhance quality degranulation (reconstruction) process, we augment by introducing vector fuzzification factors (fuzzification factor vector) and setting up an adjustment modify prototypes partition matrix. design regarded optimization problem, which guided reconstruction criterion. proposed scheme, initial matrix are generated FCM. Then introduced form appropriate for each cluster build scheme modifying With supervised mode granulation–degranulation​ composite objective function vector, Subsequently, particle swarm employed optimize refine develop optimal Finally, performance FCM algorithm enhanced. Overall, show that enhanced version process beneficial reduce deterioration results improve granulation–degranulation, also meaningful transforming data between numeric granular format. We offer thorough analysis developed scheme. particular, classical forms special case Experiments completed both synthetic publicly available datasets demonstrate approach outperforms generic approach.