trainSupv {RecordLinkage}

Train a Classifier

Description

Trains a classifier for supervised classification of record pairs.

Usage

trainSupv(rpairs, method, use.pred = FALSE, omit.possible = TRUE, 
  convert.na = TRUE, include.data = FALSE, ...)

Arguments

rpairs

Object of class RecLinkData. Training data.

method

A character vector. The classification method to use.

use.pred

Logical. Whether to use results of an unsupervised classification instead of true matching status.

omit.possible

Logical. Whether to remove pairs labeled as possible links or with unknown status.

convert.na

Logical. Whether to convert NAs to 0 in the comparison patterns.

include.data

Logical. Whether to include training data in the result object.

...

Further arguments to the training method.

Details

The given dataset is used as training data for a supervised classification. Either the true matching status has to be known for a sufficient number of data pairs or the data must have been classified previously, e.g. by using emClassify or classifyUnsup. In the latter case, argument use.pred has to be set to TRUE.

A classifying method has to be provided as a character string (factors are converted to character) through argument method. The supported classifiers are:

"svm"

Support vector machine, see svm.

"rpart"

Recursive partitioning tree, see rpart.

"ada"

Stochastic boosting model, see ada.

"bagging"

Bagging with classification trees, see bagging.

"nnet"

Single-hidden-layer neural network, see nnet.

"bumping"

A bootstrap based method using classification trees, see details.

Arguments in ... are passed to the corresponding function.

Most classifiers cannot handle NAs in the data, so by default these are converted to 0 before training.

By omit.possible = TRUE, possible links or pairs with unknown status are excluded from the training set. Setting this argument to FALSE allows three-class-classification (links, non-links and possible links), but the results tend to be poor.

Leaving include.data=FALSE saves memory, setting it to TRUE can be useful for saving the classificator while keeping track of the underlying training data.

Bumping, (acronym for “Bootstrap umbrella of model parameters”), is an ensemble method described by Tibshirani and Knight, 1999. Such as in bagging, multiple classifiers are trained on bootstrap samples of the training set. The key difference is that not the aggregated decision of all classifiers (e.g. by majority vote) is used to classify new data, but only the single model that performs best on the whole training set. In combination with classification trees as underlying classifiers this approach allows good interpretability of the trained model while being more stable against outliers than traditionally induced decision trees. The number of bootstrap samples to use can be controlled by supplying the argument n.bootstrap, which defaults to 25.

Value

An object of class RecLinkClassif with the following components:

train

If include.data is TRUE, a copy of rpairs, otherwise an empty data frame with the same column names.

model

The model returned by the underlying training function.

method

A copy of the argument method.

Author(s)

Andreas Borg, Murat Sariyar

References

Tibshirani R, Knight K: Model search by bootstrap “bumping”. Journal of Computational and Graphical Statistics 8(1999):671–686.

See Also

classifySupv for classifying with the trained model, classifyUnsup for unsupervised classification

Examples

# Train a rpart decision tree with additional parameter minsplit
data(RLdata500)
pairs=compare.dedup(RLdata500, identity=identity.RLdata500,
                    blockfld=list(1,3,5,6,7))
model=trainSupv(pairs, method="rpart", minsplit=5)
summary(model)