as.supervised {randomUniformForest}

Conversion of an unsupervised model into a supervised one

Description

Turn an unsupervised object of class unsupervised into a supervised one of class RandomUniformForest, allowing prediction of next unlabelled datasets, full analysis of variable importance in the unsupervised case and incremental unsupervised learning.

Usage

as.supervised(object, X, ...)

Arguments

object

an object of class unsupervised

X

the dataset previously learned in the unsupervised mode.

...

allow all options of randomUniformForest to be passed to the as.supervised() function, e.g. ntree, mtry, nodesize, ...

Details

The function get clusters labels and send them to the randomUniformForest classifier in order to be learnt with the data. The resulting object can be used for any coming dataset for fast prediction or clustering. Note that main argument of the method is that the model which generates the unsupervised model also generates the supervised one. The process to clustering, in both unsupervised and supervised cases, is filtered by the dissimilarity matrix in conjunction with the MDS function.

Value

An object of class randomUniformForest. See randomUniformForest.

Author(s)

Saip Ciss [e-mail]

See Also

modifyClusters, mergeClusters clusteringObservations

Examples

# see unsupervised.randomUniformForest() function.