Fuzzy density based clustering method: Soft DBSCAN-GM

The problem of unsupervised learning is that of trying to find hidden structure in unlabeled data. The last ones confide in, among other things, the choice of the clustering technique. Almost all of the well-known clustering algorithms require input number of clusters which is hard to determine but have a significant influence on the clustering result. Furthermore, the majority is not robust enough towards noisy data. In contrast, density based method DBSCAN-GM has obvious advantages over explicit samples. It discovers the number of clusters, as well as, it detects noises. Additionally, the shape of such clusters can also be irregular. However, an additional significant issue is that objects are often doubtfully specified. This vagueness may be caused by overlapping of the data regions, where one point can belong to more than one cluster. The exploit of the soft computing techniques to build groups is foreseeable in this case. For this reason, in this paper, we present an improvement of our previously defined crisp DBSCAN-GM to deal with soft objects. We name this clustering technique Soft DBSCAN-GM (SDG) that combines DBSCAN-GM and fuzzy set theory. Simulation experiments are carried out on a variety of datasets, which highlight the SDG' s effectiveness and to check the good quality of clustering results.