Seminar Notice
Statistical Laboratory
Iowa State University
DATE AND TIME: Monday, November 7, 2005, 4:10 p.m.
PLACE: 319 Snedecor
SPEAKER: Jim Booth, Cornell University, Ithaca, NY
TITLE: Clustering Using Objective Functions and Stochastic
Search
ABSTRACT
A new approach to clustering multivariate data, based on a multi-level
linear mixed model, is proposed. A key component of the model is that
observations from the same cluster are correlated, because they share cluster
specific random effects. The inclusion of such random effects allows
parsimonious deviation of the mean profile, for a given cluster, from a given
base model, that may be captured statistically via the posterior expectation, or
best linear unbiased predictor. One of the parameters in the model is the
true, underlying partition of the data, and the posterior distribution of this
parameter, which is known up to a normalizing constant, is used to cluster the
data. The problem of finding good partitions is not amenable to
deterministic methods such as the EM algorithm. Thus, we propose a simple,
Metropolis-Hastings Markov chain stochastic search, using a biased random walk
to select candidate moves. The proposed methodology is fundamentally
different from the well-known finite mixture model approach to clustering, which
requires an independent and identically distributed structure, and does not
explicitly include the partition as a parameter.
COFFEE: 3:45 p.m., 104 Snedecor Hall
Jeanette La Grange
Department of Statistics
102 Snedecor
Iowa State University
Ames, IA 50010-1210
515 294-3440 (office)
515 294-4040 (fax)