Soft Clustering Techniques: An In-Depth Analysis of GMM and FCM Algorithms and Comparative Performance

Clustering is one of the modern techniques that have been discovered to solve the problem of the degree of similarity and dissimilarity between data within the network. Clustering originates from unsupervised techniques whose main function is to organize data into subsets based on the degree of similarity between these data. The research conducted an analytical study on Fuzzy C-Means (FCM) and Gaussian Mixture Model (GMM), which are considered the most prominent clustering techniques and aims to compare them in terms of the time taken by each algorithm to cluster the data and the energy consumed. Experiments were conducted in four different scenarios. The experiments concluded that GMM showed variation in energy consumption when the number of clusters gradually increased, while FCM showed clear stability in most cases. In terms of time, GMM was generally faster with fluctuations in performance, while FCM's performance was stable but relatively slower. Ultimately, each algorithm is used in a specific environment. GMM is fast with fluctuations in performance, which is useful in applications that require speed in performance, unlike FCM, which is relatively stable but slower, which is useful in applications that require accuracy in results at the expense of time.