Multivariate location and scatter models
In this chapter we first introduce and describe different symmetrical and asymmetrical parametric and semiparametric (linear) models which are then later used as the model assumptions in the statistical analysis. The models discussed include multivariate normal distribution N p (μ,Σ) and its different extensions including multivariate t distribution t v,p (μ,Σ) distribution, multivariate elliptical distribution (μ,Σ,ρ) as well as still wider semiparametric symmetrical models. Also some models with skew distributions (generalized elliptical model, mixture models, skew-elliptical model, independent component model) are briefly discussed.
KeywordsIndependent Component Analysis Independent Component Analysis Multivariate Normal Distribution Scatter Matrix Scatter Model
Unable to display preview. Download preview PDF.