prtAction - Base class for many PRT components.
prtAction is an abstract class and cannot be instantiated.
Classification, regression and feature selection techniques are all
sub-classes of prtAction.
All prtAction objects have the following properties:
name - Descriptive name for prtAction object
nameAbbreviation - Shortened name for prtAction object
isTrained - Indicates whether the current prtAction
object has been trained
isCrossValidateValid - Flag indicating whether or not
cross-validation is a valid operation on
this prtAction object
verboseStorage - Flag to allow or disallow verbose storage
dataSetSummary - A struct, set during training, containing
information about the training data set
dataSet - A prtDataSet, containing the training data,
only used if verboseStorage is true
userData - A struct containing user-specified data
All prtAction objects have the following methods:
train - Train the prtAction object using a prtDataSet
run - Evaluate the prtAction object on a prtDataSet
runOnTrainingData - Evaluate the prtAction object on a prtDataSet
during training prior to training of other
prtActions within a prtAlgorithm
crossValidate - Cross-validate the prtAction object using a
labeled prtDataSet and cross-validation keys
kfolds - K-folds cross-validate the prtAction object
using a labeled prtDataSet
optimize - Optimize the prtAction for a specified
parameter