spikeSlabGAM_0.9-6 (initial public release)
spikeSlabGAM implements Bayesian variable selection, model choice,
and regularized estimation in (geo-)additive mixed models for
Gaussian, binomial, and Poisson responses.
Its purpose is (1) to choose an appropriate subset of potential
covariates and their interactions, (2) to determine whether linear
or more flexible functional forms (P-splines, tensor product
splines) are required to model the (joint) effects of the respective
covariates, and (3) to fit these regularized effects and return
(model-averaged) estimates.
Selection and regularization of the model terms is based on a novel
spike-and-slab-type prior on coefficient groups associated with
parametric and semi-parametric effects. Inference is fully Bayesian
with an underlying MCMC sampler implemented in C and can take
advantage of multi-core processors via multicore or snow. The
package uses standard formula syntax so that complex models can be
specified very concisely. It features powerful and user friendly
visualizations using ggplot2.
----------------------------------------------------------------------
Fabian Scheipl
Department of Statistics
Ludwig-Maximilians-University Munich
Ludwigstr. 33, room 239
80539 Munich
Germany
Phone: +49-89-2180-2284
http://www.statistik.lmu.de/~scheipl/
_______________________________________________
R-packages mailing list
R-packages at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-packages
