mixvarselect: Mixture of expert modelling with variable selection using penalization

Share:

The package uses Ridge, Lasso and SCAD penlization to implement variable selection and regularization in various parameters of mixutres of Gaussians. The package offers variable selection for the mean compoenents and logit of mixture proportions. The package requires installation of GSL libraries on the machine and may not work under Windows platform.

Author
Vahid Partovi Nia and Abbas Khalili
Date of publication
2015-06-06 23:19:47
Maintainer
Vahid Partovi Nia <vahid.partovinia@polymtl.ca>
License
GPL (>= 2)
Version
1.1

View on R-Forge

Man pages

moeridge
mixture of expert fit with ridge penalization
moesim
Simulated data

Files in this package

mixvarselect/DESCRIPTION
mixvarselect/NAMESPACE
mixvarselect/R
mixvarselect/R/moe.R
mixvarselect/R/zzz.R
mixvarselect/data
mixvarselect/data/moesim.rda
mixvarselect/man
mixvarselect/man/moeridge.Rd
mixvarselect/man/moesim.Rd
mixvarselect/src
mixvarselect/src/rmoe-adapted.c