The bivariate $K$-finite normal mixture "blanket" copula: an application to driving patterns

摘要: There are many bivariate parametric copulas in the literature to model data with different dependence features. We propose a new copula family that cannot only handle various patterns appear existing families, but also provides more enriched structure. The proposed construction exploits finite mixtures of normal distributions. mixing operation, distinct correlation and mean parameters at each mixture component introduce quite flexible dependence. apply real transportation apparently be modelled by any families copulas.