A new construction for skew multivariate distributions (Q2567123): Difference between revisions

Starting with an elliptically contoured multivariate distribution and then imposing linear constraints on the vector random variable, the authors arrive at a general class of distributions. The authors call them skew elliptic distributions. By considering different types of contraints they obtain several classes of skew multivariate distributions. It is shown that the skew distributions obtained by other authors are special cases of the new distributions. Moment generating functions for these classes of skew multivariate distributions are also derived.