Estimation for Dirichlet Mixed Models
Abstract
Dirichlet mixed models find wide application and the estimation of these models is usually achieved through the method of moments. We present an iterative hybrid algorithm for obtaining the (MLE) maximum likelihood estimate employing both modified Newton-Raphson and E-M (electro-magnetics) methods. This successful MLE algorithm enables calculation of a jackknife MLE. Simulation comparison of the three estimates is provided. The MLE substantially improves upon the moments estimator particularly with increasing dimension and the jackknife MLE in turn offers a more dramatic improvement. Keywords: Comparison of estimates, Mixed moments, Moment estimator.
Document Details
- Document Type
- Technical Report
- Publication Date
- May 19, 1990
- Accession Number
- ADA223583
Entities
People
- Alan E. Gelfand
- Steve Leeds
Organizations
- Stanford University