Skip to main content
Fig. 1 | Environmental Microbiome

Fig. 1

From: Reducing the arbitrary: fuzzy detection of microbial ecotones and ecosystems – focus on the pelagic environment

Fig. 1

a Typical outputs of K-means clustering (2 clusters). The separation between the two clusters is crisp, as observations can only belong to one cluster, regardless of their actual distance to the cluster centroids. b Typical outputs of fuzzy clustering (same 2 clusters). Each observation now belongs to both clusters, according to their degree of similarity with each cluster centroid. This translates in their membership grades (e.g. 65% in cluster 1 and 35% in cluster 2). There is no strict boundary between the two clusters, as they now overlap. c Schematic representation of the membership grade profile of a single cluster over the spatial dimensions of the sampling. d Schematic representation of the overlap between the two clusters membership grade profiles over the spatial dimensions of the sampling

Back to article page