INVESTIGADORES
CARANDO Daniel German
artículos
Título:
Nonparametric likelihood based estimation for a multivariate Lipschitz density
Autor/es:
DANIEL CARANDO; RICARDO FRAIMAN; PABLO GROISMAN
Revista:
JOURNAL OF MULTIVARIATE ANALYSIS
Editorial:
ELSEVIER INC
Referencias:
Año: 2009 vol. 100 p. 981 - 992
ISSN:
0047-259X
Resumen:
We consider a problem of nonparametric density estimation undershape restrictions. We deal with the case where the density belongsto a class of Lipschitz functions. Devroye (1987) considered theseclasses of estimates as tailor-made estimates, in opposition in someway to universally consistent estimates. In our framework we getexistence and uniqueness of the maximum likelihood estimate as wellas strong consistency. This NPMLE can be easily characterized but itis not easy to compute. Some simpler approximations are alsoconsidered.
