Define the default configuration parameters for the DependencyBasedStrategy strategy.
D2MCS::StrategyConfiguration -> DependencyBasedStrategyConfiguration
new()Method for initializing the object arguments during runtime.
DependencyBasedStrategyConfiguration$new(
binaryCutoff = 0.6,
realCutoff = 0.6,
tiebreakMethod = "lfdc",
metric = "dep.tar"
)binaryCutoffThe numeric value of binary cutoff.
realCutoffThe numeric value of real cutoff.
tiebreakMethodThe character value of tie-break method. The two tiebreak methods available are "lfdc" (less dependence cluster with the features) and "ltdc" (less dependence cluster with the target). These methods are used to add the features in the candidate feature clusters.
metricThe character value of the metric to apply the mean to obtain the quality of a cluster. The two metrics available are "dep.tar" (Dependence of cluster features on the target) and "dep.fea" (Dependence between cluster features).
minNumClusters()Function used to return the minimum number of clusters distributions used. By default the minimum is set in 2.
DependencyBasedStrategyConfiguration$minNumClusters(...)...Further arguments passed down to minNumClusters
function.
A numeric vector of length 1.
maxNumClusters()The function is responsible of returning the maximum number of cluster distributions used. By default the maximum number is set in 50.
DependencyBasedStrategyConfiguration$maxNumClusters(...)...Further arguments passed down to maxNumClusters
function.
A numeric vector of length 1.
getBinaryCutoff()Gets the cutoff to consider the dependency between binary features.
DependencyBasedStrategyConfiguration$getBinaryCutoff()The numeric value of binary cutoff.
getRealCutoff()Gets the cutoff to consider the dependency between real features.
DependencyBasedStrategyConfiguration$getRealCutoff()The numeric value of real cutoff.
setBinaryCutoff()Sets the cutoff to consider the dependency between binary features.
DependencyBasedStrategyConfiguration$setBinaryCutoff(cutoff)cutoffThe new numeric value of binary cutoff.
setRealCutoff()Sets the cutoff to consider the dependency between real features.
DependencyBasedStrategyConfiguration$setRealCutoff(cutoff)cutoffThe new numeric value of real cutoff.
tiebreak()The function solves the ties between two (or more) features.
DependencyBasedStrategyConfiguration$tiebreak(
feature,
clus.candidates,
fea.dep.dist.clus,
corpus,
heuristic,
class,
class.name
)featureA character containing the name of the feature
clus.candidatesA single or numeric vector value to identify the candidate groups to insert the feature.
fea.dep.dist.clusA list containing the groups chosen for the features.
corpusA data.frame containing the features of the initial data.
heuristicThe heuristic used to compute the relevance of each feature. Must inherit from GenericHeuristic abstract class.
classA character vector containing all the values of the target class.
class.nameA character value representing the name of the target class.
qualityOfCluster()The function determines the quality of a cluster.
DependencyBasedStrategyConfiguration$qualityOfCluster(clusters, metrics)clustersA list with the feature distribution of each cluster.
metricsA numeric list with the metrics associated to the cluster (dependency between all features and dependency between the features and the class).
A numeric vector of length 1.
isImprovingClustering()The function indicates if clustering is getting better as the number of them increases.
DependencyBasedStrategyConfiguration$isImprovingClustering(clusters.deltha)clusters.delthaA numeric vector value with the quality values of the built clusters.
A numeric vector of length 1.
clone()The objects of this class are cloneable with this method.
DependencyBasedStrategyConfiguration$clone(deep = FALSE)deepWhether to make a deep clone.
StrategyConfiguration,
DependencyBasedStrategy