A0438
Title: A semiparametric pairwise likelihood for mixed data in copula models
Authors: Ekaterina Tomilina - INRAE (France)
Florence Jaffrezic - INRAE (France)
Gildas Mazo - INRA (France) [presenting]
Abstract: Jointly analyzing mixed-type data in high dimensions is a difficult problem, and yet many applications fall into this case. A semiparametric pairwise likelihood method to estimate the parameters in copula models from a sample of independent and identically distributed observations of mixed-type variables is derived and studied theoretically and numerically. Applications in genomics are presented with the Gaussian copula.
