The Method of Sieves. A Survey of Recent Applications.
FLORIDA STATE UNIV TALLAHASSEE DEPT OF STATISTICS
Pagination or Media Count:
The method of sieves is a technique of nonparametric estimation in which estimators are restricted by an increasing sequence of subsets of the parameter space with the subsets indexed by the sample size. The need for this technique arises in situations where the parameter space is too large for the existence or consistency of unconstrained maximum likelihood or least squares estimators. Grenander 10 developed the abstract theory of the method of sieves and provided a wealth of examples illustrating its use. This article is to appear under the entry SIEVES, METHOD OF in the Encylopedia of Statistical Sciences.
- Statistics and Probability