sda-package {sda} | R Documentation |
This package performs linear and diagonal discriminant analysis with variable selection.
The classifier is trained using James-Stein-type shrinkage estimators. Variable selection is based on a univariate test statistics (multi-class cat score) and subsequent false non-discovery rate (FNDR) testing. In addition, higher criticism (HC) feature scores are computed and can also be used for feature thresholding.
This approach is particularly suited for high-dimensional classification. For details see Ahdesm"aki and Strimmer (2009).
Typically the functions in this package are applied in three steps:
sda.ranking
,sda
, andpredict.sda
.The accompanying web site (see below) provides example R scripts to illustrate the functionality of this package.
Miika Ahdesm"aki and Korbinian Strimmer (http://strimmerlab.org/)
See website: http://strimmerlab.org/software/sda/
sda.ranking
, sda
, predict.sda
.