An S4 class representing the evaluation of a learning method by repeated sampling.
Constructor for the S4 Renoir object.
Usage
Renoir(
id = character(),
config = character(),
response = character(),
nout = integer(),
grid = integer(),
k = integer(),
sampling = character(),
learning = character(),
screening = character(),
scoring = character(),
filter = Filtered(),
evaluation = EvaluatedList(),
stability = list(),
marks = MarkedList(),
nbest = data.frame(),
call = list()
)Arguments
- id
evaluated learning method
- config
name of the configuration
- response
response type
- nout
the number of responses
- grid
integer vector, the considered training set sizes
- k
integer, the number of sampling repeats
- sampling
string, sampling method used for evaluation
- learning
string, the learning procedure (i.e. training or tuning)
- screening
string, the feature screening strategy adopted
- scoring
string, the accuracy measure(s) used during the evaluation
- filter
a Filtered object containing the summary of the pre-processing
- evaluation
a EvaluatedList object containing the evaluation of the learning method
- stability
list, containing the features stability for each response, i.e. the frequency of features recruitment per response across all the computed models
- marks
a MarkedList object containing the features importance
- nbest
data.frame, it contains the indices of the automatically selected best training set size for different settings (e.g. assessment on training, testing, full set of data)
- call
list, containing the arguments to the "renoir" function call that creates the Renoir object
Slots
ida character string
configa character string
responsea character string
nouta one-length integer vector
grida integer vector
ka one-length integer vector
samplinga character string
learninga character string
screeninga character string
scoringa character string
filtera Filtered object
evaluationa EvaluatedList object
stabilitya list object
marksa MarkedList object
nbesta data.frame object
calllist