Accelerated Fuzzy C-Means Clustering Based on New Affinity Filtering and Membership Scaling

Fuzzy C-Means (FCM) is a widely used clustering method. However, FCM and its many accelerated variants have low efficiency in the mid-to-late stage of process. In this stage, all samples are involved updating their non-affinity centers, fuzzy membership grades most samples, whose assignments remain unchanged, still updated by calculating sample-center distances. All these factors lead to algorithms converging slowly. paper, new affinity filtering technique developed recognize complete set centers for each sample with computations. Then, scaling suggested between 0 maintain others. By integrating two techniques, based on (AMFCM) proposed accelerate whole convergence process FCM. Numerous experimental results performed synthetic real-world data sets shown feasibility algorithm. Compared state-of-the-art algorithms, AMFCM significantly faster more effective. For example, reduces number iterations 80 $\%$ average.