Parameter expansion for estimation of reduced rank covariance matrices(Open Access publication)
Author:
Open Access
- 21 December 2007
- journal article
- review article
- Published by Springer Nature in Genetics Selection Evolution
- Vol. 40 (1) , 3-24
Abstract
Parameter expanded and standard expectation maximisation algorithms are described for reduced rank estimation of covariance matrices by restricted maximum likelihood, fitting the leading principal components only. Convergence behaviour of these algorithms is examined for several examples and contrasted to that of the average information algorithm, and implications for practical analyses are discussed. It is shown that expectation maximisation type algorithms are readily adapted to reduced rank estimation and converge reliably. However, as is well known for the full rank case, the convergence is linear and thus slow. Hence, these algorithms are most useful in combination with the quadratically convergent average information algorithm, in particular in the initial stages of an iterative solution scheme.Keywords
This publication has 38 references indexed in Scilit:
- PERSPECTIVES OF ANOVA, REML AND A GENERAL LINEAR MIXED MODEL2004
- The PX-EM algorithm for fast stable fitting of Henderson's mixed modelGenetics Selection Evolution, 2000
- Acceleration of the EM Algorithm by using Quasi-Newton MethodsJournal of the Royal Statistical Society Series B: Statistical Methodology, 1997
- Matrix Algebra From a Statistician’s Perspective1997
- Numerical Methods for Unconstrained Optimization and Nonlinear Equations1996
- Computing Modified Newton Directions Using a Partial Cholesky FactorizationSIAM Journal on Scientific Computing, 1995
- A reparameterization to improve numerical optimization in multivariate REML (co)variance component estimationGenetics Selection Evolution, 1994
- Estimation of Variances and Covariances under Multiple Trait ModelsJournal of Dairy Science, 1984
- Maximum Likelihood Approaches to Variance Component Estimation and to Related ProblemsJournal of the American Statistical Association, 1977
- A Simple Method for Computing the Inverse of a Numerator Relationship Matrix Used in Prediction of Breeding ValuesBiometrics, 1976