combineUnsupervised: Combine Unsupervised Learning objects

In randomUniformForest: Random Uniform Forests for Classification, Regression and Unsupervised Learning

Combine Unsupervised Learning objects

Description

Combine unsupervised learning objects in order to achieve incremental learning. Only the MDS (spectral) points are used before calling a clustering algorithm on all. Note that the function is currently highly experimental with a lack of applications.

Usage

combineUnsupervised(...)

Arguments

...	(enumeration of) objects of class unsupervised, coming from unsupervised.randomUniformForest, that needs to be combined.

Value

An object of class unsupervised, which is a list with the following components:

proximityMatrix	the resulted dissimilarity matrix.
MDSModel	the resulted Multidimensional scaling model.
unsupervisedModel	the resulted unsupervised model with clustered observations in unsupervisedModel$cluster.
largeDataLearningModel	if the dataset is large, the resulted model that learned a sample of the MDS points, and predicted others points.
gapStatistics	if K-means algorithm has been called, the results of the gap statistic. Otherwise NULL.
rUFObject	Random Uniform Forests object.
nbClusters	Number of clusters found.
params	options of the model.

Author(s)

Saip Ciss [e-mail]

See Also

update.unsupervised, modifyClusters, mergeClusters, splitClusters, clusteringObservations, as.supervised

Examples

## not run
## Wine Quality Data Set
## http://archive.ics.uci.edu/ml/datasets/Wine+Quality

# data(wineQualityRed)
# X = wineQualityRed[, -ncol(wineQualityRed)]

## 1 - run unsupervised analysis on the first half of dataset 

# subset.1 = 1:floor(nrow(X)/2)
# wineQualityRed.model.1 = unsupervised.randomUniformForest(X, subset = subset.1, depth = 5) 

## assess roughly the model and visualize
#  wineQualityRed.model.1

# plot(wineQualityRed.model.1)

## 2 - run unsupervised analysis on the second half of dataset
# wineQualityRed.model.2 = unsupervised.randomUniformForest(X, subset = -subset.1, depth = 5)


## 2.1 if less clusters (than in 1) are got, split the one with the highest number of cases
## it is the second cluster in our case
# wineQualityRed.model.2 = splitClusters(wineQualityRed.model.2, 2)

## roughly assess and, eventually, merge and split again (with different seeds) in order 
## to be confident about the new clustering
# wineQualityRed.model.2

## 3 - combine
# wineQualityRed.combinedModel = 
# combineUnsupervised(wineQualityRed.model.1, wineQualityRed.model.2)

## visualize and plot
# wineQualityRed.combinedModel
# plot(wineQualityRed.combinedModel)

## compare with the full data and same modelling
# wineQualityRed.model = unsupervised.randomUniformForest(X, depth = 5)

## or increase depth (more computation and default option) for a more detailed model
# wineQualityRed.model = unsupervised.randomUniformForest(X)

randomUniformForest documentation built on June 22, 2022, 1:05 a.m.