Using SOM as a Tool for Automated Design of Clustering Systems Based on Fuzzy Predicates

Automatic design of interpretable fuzzy predicate systems for clustering using self-organizing maps

“…We tested the method using the following datasets: Iris data (3 classes, 4 features, 150 data) [20]. Wine data (3 classes, 13 features, 198 data) [41].…”

“…Unlike previous works, in this paper we propose the design of a new clustering system in which (a) we use predicate fuzzy logic [19] to perform the clustering task, which is a natural extension of predicate Boolean Logic and (b) the system is automatically designed (unsupervised) [20] using a two-level clustering approach that combines SOM and Fuzzy C-Means (FCM) as a second level clustering method. First, SOM are trained from the original data, considering a SOM with a number of cells much larger than the expected number of clusters.…”

“…Next, the codebook clustering is analyzed and membership functions and predicates are defined. Thus we obtain a ranked clustering criteria represented as selfdiscovered fuzzy predicates using data information [20] which consider the behavior of the variables into the different clusters. Hereafter we will name the proposed method SOM-based Fuzzy Predicate Clustering (SFPC).…”

Interpretable interval type-2 fuzzy predicates for data clustering: A new automatic generation method based on self-organizing maps

“…FL has been applied to data clustering and classification applications both using experts' knowledge to generate fuzzy models as well as proposing automatic methods based on data analysis. The papers [2,3,14,15] are examples of such applications and, also, the surveys [7,11] .…”

“…When applied on data clustering, fuzzy predicate models allow to implement knowledge about the clustering, explaining which values of each feature are related to each of the clusters and modeling these relationships using membership functions and predicates. Such models have been successfully applied in data clustering [2,3,14,15] , having the following features:…”