trainTuneParetoClassifier: Train a TunePareto classifier
In TunePareto: Multi-Objective Parameter Tuning for Classifiers

trainTuneParetoClassifier

R Documentation

Train a TunePareto classifier

Description

Trains a classifier wrapped in a TuneParetoClassifier object. The trained classifier model can then be passed to predict.TuneParetoModel.

Usage

trainTuneParetoClassifier(classifier, trainData, trainLabels, ...)

Arguments

`classifier`	A `TuneParetoClassifier` object as returned by `tuneParetoClassifier` or one of the predefined classification functions (see `predefinedClassifiers`).
`trainData`	The data set to be used for the classifier training. This is usually a matrix or data frame with the samples in the rows and the features in the columns.
`trainLabels`	A vector of class labels for the samples in `trainData`.
`...`	Further parameters to be passed to the classifier. These must be parameters specified in the `classifierParameterNames` parameter of `tuneParetoClassifier` and usually correspond to the tuned parameters.

Value

Returns an object of class TuneParetoModel with the following entries

`classifier`	The classifier object supplied in the `classifier` parameter
`classifierParams`	The additional parameters supplied to the classifier in the `...` parameter
`trainData`	If `classifier` is an all-in-one classifier without a separate prediction method, this stores the input training data.
`trainLabels`	If `classifier` is an all-in-one classifier without a separate prediction method, this stores the input training labels.
`model`	If `classifier` consists of separate training and prediction methods, this contains the trained classifier model.

Examples


# train an SVM classifier
cl <- tunePareto.svm()
tr <- trainTuneParetoClassifier(cl, 
                                iris[,-ncol(iris)], 
                                iris[,ncol(iris)], 
                                cost=0.001)

# re-apply the classifier to predict the training data
print(iris[,ncol(iris)])
print(predict(tr, iris[,-ncol(iris)]))

TunePareto documentation built on Oct. 2, 2023, 5:06 p.m.