Clustering in linear mixed models with approximate Dirichlet process mixtures using EM algorithm Mathematik, Informatik und Statistik - Open Access LMU - Teil 02/03

    • Education

In linear mixed models, the assumption of normally distributed random effects is often inappropriate and unnecessarily restrictive. The proposed approximate Dirichlet process mixture assumes a hierarchical Gaussian mixture that is based on the truncated version of the stick breaking presentation of the Dirichlet process. In addition to the weakening of distributional assumptions, the specification allows to identify clusters of observations with a similar random effects structure. An Expectation-Maximization algorithm is given that solves the estimation problem and that, in certain respects, may exhibit advantages over Markov chain Monte Carlo approaches when modelling with Dirichlet processes. The method is evaluated in a simulation study and applied to the dynamics of unemployment in Germany as well as lung function growth data.

In linear mixed models, the assumption of normally distributed random effects is often inappropriate and unnecessarily restrictive. The proposed approximate Dirichlet process mixture assumes a hierarchical Gaussian mixture that is based on the truncated version of the stick breaking presentation of the Dirichlet process. In addition to the weakening of distributional assumptions, the specification allows to identify clusters of observations with a similar random effects structure. An Expectation-Maximization algorithm is given that solves the estimation problem and that, in certain respects, may exhibit advantages over Markov chain Monte Carlo approaches when modelling with Dirichlet processes. The method is evaluated in a simulation study and applied to the dynamics of unemployment in Germany as well as lung function growth data.

Top Podcasts In Education

The Mel Robbins Podcast
Mel Robbins
The Jordan B. Peterson Podcast
Dr. Jordan B. Peterson
TED Talks Daily
TED
UNBIASED
Jordan Is My Lawyer
Do The Work
Do The Work
The Rich Roll Podcast
Rich Roll

More by Ludwig-Maximilians-Universität München

Center for Advanced Studies (CAS) Cutting Edge (LMU) - SD
Center for Advanced Studies (CAS)
Hegel lectures by Robert Brandom, LMU Munich
Robert Brandom, Axel Hutter
Medizinische Fakultät - Digitale Hochschulschriften der LMU - Teil 04/19
Ludwig-Maximilians-Universität München
Volkswirtschaft - Open Access LMU - Teil 01/03
Ludwig-Maximilians-Universität München
Volkswirtschaftliche Fakultät - Digitale Hochschulschriften der LMU
Ludwig-Maximilians-Universität München
Women Thinkers in Antiquity and the Middle Ages - SD
Peter Adamson