Flexible Clustering for High-Dimensional Data via Mixtures of Joint Generalized Hyperbolic Models
Methodology
2018-11-02 v2 Computation
Abstract
A mixture of joint generalized hyperbolic distributions (MJGHD) is introduced for asymmetric clustering for high-dimensional data. The MJGHD approach takes into account the cluster-specific subspace, thereby limiting the number of parameters to estimate while also facilitating visualization of results. Identifiability is discussed, and a multi-cycle ECM algorithm is outlined for parameter estimation. The MJGHD approach is illustrated on two real data sets, where the Bayesian information criterion is used for model selection.
Keywords
Cite
@article{arxiv.1705.03130,
title = {Flexible Clustering for High-Dimensional Data via Mixtures of Joint Generalized Hyperbolic Models},
author = {Yang Tang and Ryan P. Browne and Paul D. McNicholas},
journal= {arXiv preprint arXiv:1705.03130},
year = {2018}
}