Finite mixtures of von Mises-Fisher distributions allow to apply model-based clustering methods to data which is of standardized length, i.e., all data points lie on the unit sphere. The R package movMF contains functionality to draw samples from finite mixtures of von
Mises-Fisher distributions and to fit these models using the expectation-maximization algorithm
for maximum likelihood estimation. Special features are the possibility to use
sparse matrix representations for the input data, different variants of the expectationmaximization
algorithm, different methods for determining the concentration parameters
in the M-step and to impose constraints on the concentration parameters over the components.
In this paper we describe the main fitting function of the package and illustrate its
application. In addition we compare the clustering performance of finite mixtures of
von Mises-Fisher distributions to spherical k-means. We also discuss the resolution of
several numerical issues which occur for estimating the concentration parameters and for
determining the normalizing constant of the von Mises-Fisher distribution. (authors' abstract)
| Identifer | oai:union.ndltd.org:VIENNA/oai:epub.wu-wien.ac.at:4893 |
| Date | 07 1900 |
| Creators | Hornik, Kurt, Grün, Bettina |
| Publisher | American Statistical Association |
| Source Sets | Wirtschaftsuniversität Wien |
| Language | English |
| Detected Language | English |
| Type | Article, PeerReviewed |
| Format | application/pdf |
| Relation | http://dx.doi.org/ 10.18637/jss.v058.i10, http://www.jstatsoft.org/, http://epub.wu.ac.at/4893/ |
Page generated in 0.0019 seconds