CMStatistics 2018: Start Registration
View Submission - CMStatistics
Title: The fclust R package for fuzzy clustering: A new version Authors:  Maria Brigida Ferraro - Sapienza University of Rome (Italy)
Paolo Giordani - Sapienza University of Rome (Italy)
Alessio Serafini - Sapienza Universita di Roma (Italy) [presenting]
Abstract: Fuzzy clustering is a technique extensively used in several domains of research to discover fuzzy partitions of data sets. The fclust package is a useful toolbox for fuzzy clustering in R programming language. Differently from the other fuzzy clustering packages available on CRAN, fclust implements not only the most known fuzzy clustering algorithm, the fuzzy $k$-means (fkm), but also many extensions of fkm, cluster validity indices and visualization tools. The new version allows us to use fuzzy relational clustering algorithms for partitioning mixed-type data, a new version of Gustafson-Kessel algorithm to avoid singularity in the covariance matrix, fuzzy versions of some cluster comparison methods, and few minor changes as automatically selection of number of clusters.