A note on the representation of parametric models for multivariate extremes

From MaRDI portal





The author extends the definition of bivariate extremal logistic distribution to the \(d\)-dimensional case. A random variable \(W=(W_1,\dots,W_d)\) on the standard simplex \(S_d\) in \(\mathbb R^d\) has extremal logistic distribution (ELD) with parameters \(\alpha_1\),\dots, \(\alpha_d\) if it is a solution to \[ {W_j\over W_d}={C_j U_j^{-\alpha_j}\over C_d U_d^{-\alpha_d}},\;j=1,\dots, d, \] where \(C_j=\Gamma(d-\alpha_j)/\Gamma(1-\alpha_j)\), \(U=(U_1,\dots,U_d)\) has density \[ h_U(u)={1\over d}\sum_{j=1}^d {\Gamma(d-\alpha_j)\over\Gamma(1-\alpha_j)} u_j^{-\alpha_j}. \] Some representations of ELDs and the extreme Dirichlet distribution are discussed. A simulation algorithm is proposed based on these representations.





Describes a project that uses

Uses Software






This page was built for publication: A note on the representation of parametric models for multivariate extremes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q626291)