Asymmetric Clustering for High-Dimensional Data via Mixtures of Joint Generalized Hyperbolic Models

A mixture of joint generalized hyperbolic models is introduced for asymmetric clustering for high-dimensional data (MJGHM-HDClust). The MJGHM-HDClust approach takes into account the cluster specific subspace and therefore limits the number of parameters to estimate. Parameter estimation is carried using the multi-cycle ECM algorithm and the Bayesian information criterion is used for model selection. Real and simulated data are used to demonstrate the MJGHM-HDClust approach.