Robust variable selection for REDDA models via TBIC approximation to the integrated likelihood. It is a robust adaptation of the greedy forward search algorithm for discriminant analysis. Based on clustvarsel R package.
This repository is associated with the papers
- Cappozzo, Greselin, Murphy (2021) Robust variable selection for model-based learning in presence of adulteration. https://doi.org/10.1016/j.csda.2021.107186
- Cappozzo, Duponchel, Greselin, Murphy (2021) Robust variable selection in the framework of classification with label noise and outliers: Applications to spectroscopic data in agri-food. https://doi.org/10.1016/j.aca.2021.338245
You can install the development version of varselTBIC from GitHub with:
# install.packages("devtools")
devtools::install_github("AndreaCappozzo/varselTBIC")