HDclassif: AnRPackage for Model-Based Clustering and Discriminant Analysis of High-Dimensional Data

Abstract

This paper presents the R package HDclassif which is devoted to the clustering and the discriminant analysis of high-dimensional data. The classification methods proposed in the package result from a new parametrization of the Gaussian mixture model which combines the idea of dimension reduction and model constraints on the covariance matrices. The supervised classification method using this parametrization is called high dimensional discriminant analysis (HDDA). In a similar manner, the associated clustering method is called high dimensional data clustering (HDDC) and uses the expectation-maximization algorithm for inference. In order to correctly fit the data, both methods estimate the specific subspace and the intrinsic dimension of the groups. Due to the constraints on the covariance matrices, the number of parameters to estimate is significantly lower than other model-based methods and this allows the methods to be stable and efficient in high dimensions. Two introductory examples illustrated with R codes allow the user to discover the hdda and hddc functions. Experiments on simulated and real datasets also compare HDDC and HDDA with existing classification methods on high-dimensional datasets. HDclassif is a free software and distributed under the general public license, as part of the R software project.

Model-based clustering of high-dimensional data: A review

“…Firstly, the package HDclassif [7] for the R software provides routines for model-based clustering and classification of high-dimensional data. In particular, the hddc function implements the subspace clustering method of [20].…”

Characterization of Neighborhood Behaviours in a Multi-neighborhood Local Search Algorithm

“…We can start doing cluster analysis on neighborhoods based on those vectors. Since the number of dimensions (120) is larger than the number of individuals (42), the clustering method High-Dimensional Data Clustering (HDDC) [12]), which is implemented in the R package HDclassif [13], is used for cluster analysis. This method has two desirable properties: the ability of dealing with high-dimensional low-sample data, and the optimal number of clusters automatically decided based on Bayesian Information Criterion.…”

Probabilistic model‐based discriminant analysis and clustering methods in chemometrics

“…We here focus on the problem of discriminant analysis and we consider the 3-class NIR data set presented in [26]. The 3-class NIR data set contains 221 NIR spectra of manufactured textiles of various compositions, the classification problem consisting in the determination of a physical property which can take three discrete values [7]. The NIR spectra were measured on a XDS rapid content analyzer instrument (FOSS) in reflectance mode in the range 1100-2500 nm at 0.5 nm apparent resolution (2800 data points per spectrum).…”

“…We consider the HDDA method which is implemented in the HDclassif package [7] for R. HDDA is the supervised classification method associated with the HD-GMM model presented in Section 4.2 and it is a subspace classification method.…”