FADA: Variable selection for supervised classification in high dimension

The functions provided in the FADA (Factor Adjusted Discriminant Analysis) package aim at performing supervised classification models and variable selection on dependent covariates. The classification procedures are combined with a factor modeling of dependence among covariates. The available procedures are Lasso regularized logistic model (see Friedman et al. (2010)), sparse linear discriminant analysis (see Clemmensen et al. (2011)), shrinkage linear and diagonal discriminant analysis (see M. Ahdesmaki et al. (2010)).

Downloads: