A0474
Title: Pivotal consensus clustering through the \texttt{pivmet} R package
Authors: Leonardo Egidi - University of Trieste (Italy) [presenting]
Roberta Pappada - University of Trieste (Italy)
Francesco Pauli - University of Trieste (Italy)
Nicola Torelli - University of Trieste (Italy)
Abstract: Identifying the prototypes of a group, i.e. the elements of a dataset representing different groups of data points, is relevant to the tasks of clustering, classification and mixture modeling. The R package pivmet provides functions to perform consensus clustering based on pivotal units, which may allow the improvement of classical techniques such as the standard k-means algorithm via careful seeding; moreover, the package is flexibly programmed to support applications to real and simulated datasets. Finally, some preliminary procedures aimed at identifying the number of clusters based on a co-association matrix will be illustrated.
