Seminar
Statistical Laboratory
Iowa State University
 
DATE AND TIME: Monday, November 14, 2005, 4:10 p.m.
 
PLACE: 319 Snedecor
 
SPEAKER:  Paul Speckman, Department of Statistics, University of Missouri-Columbia
 
TITLE: Spatially Adaptive Smoothing Splines
 
ABSTRACT
 
Smoothing splines and more generally penalized smoothing methods are popular for curve-fitting and for function estimation in additive models.  Moreover, there is a well-known interpretation showing that these methods arise as Bayesian estimators from intuitively appealing prior distributions.  One can view the smoothing problem as balancing fidelity to the data against smoothness through a
single parameter.  We show how the model can be extended to adapt to various degrees of smoothness in the data.  The model is equivalent to a spatially adaptive prior.  Two methods are presented for fitting the models, one using generalized cross validation and a second fully Bayesian approach suitable for Markov random field type priors.  Applications are made to nonparametric regression and density estimation.
 
COFFEE:  3:45 p.m., 104 Snedecor Hall
 
Seminar schedules and abstracts are available via WWW:  http://www.stat.iastate.edu/

Jeanette La Grange
Department of Statistics
102 Snedecor
Iowa State University
Ames, IA  50010-1210
515 294-3440 (office)
515 294-4040 (fax)
http://www.stat.iastate.edu/directory/staff/jeanette.html