Flexible Multivariate Density Estimation with Marginal Adaptation - Statistics > MethodologyReportar como inadecuado

Flexible Multivariate Density Estimation with Marginal Adaptation - Statistics > Methodology - Descarga este documento en PDF. Documentación en PDF para descargar gratis. Disponible también para leer online.

Abstract: Our article addresses the problem of flexibly estimating a multivariatedensity while also attempting to estimate its marginals correctly. We do so byproposing two new estimators that try to capture the best features of mixtureof normals and copula estimators while avoiding some of their weaknesses. Thefirst estimator we propose is a mixture of normals copula model that is aflexible alternative to parametric copula models such as the normal and tcopula. The second is a marginally adapted mixture of normals estimator thatimproves on the standard mixture of normals by using information contained inunivariate estimates of the marginal densities. We show empirically that copulabased approaches can behave much better or much worse than estimators based onmixture of normals depending on the properties of the data. We provide fast andreliable implementations of the estimators and illustrate the methodology onsimulated and real data.

Autor: Paolo Giordani, Xiuyan Mun, Robert Kohn

Fuente: https://arxiv.org/

Documentos relacionados