Large sample results under biased sampling when covariables are present
In this work, we introduce and analyze new estimation procedures when a (p+1)-dimensional variable (U,T) is sampled under selection bias models. Here, T denotes a time of interest and U=(u1,...,up) is a vector of covariables. Applications include the estimation of (conditional) cumulative hazard and mean residual time functions, and inference about the regression curve m(u)=E([psi](T) U=u).
Year of publication: |
2003
|
---|---|
Authors: | de Uña-Álvarez, Jacobo |
Published in: |
Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 63.2003, 3, p. 287-293
|
Publisher: |
Elsevier |
Keywords: | Covariables Least squares Regression Selection bias Weighted distributions |
Saved in:
Saved in favorites
Similar items by person
-
Non-parametric k-sample tests: Density functions vs distribution functions
Martínez-Camblor, Pablo, (2009)
-
Distributional convergence under proportional censorship when covariables are present
de Uña-Álvarez, Jacobo, (1998)
-
A simple estimator of the bivariate distribution function for censored gap times
de Uña-Álvarez, Jacobo, (2008)
- More ...