jsat.datatransform.featureselection

Class LRS

java.lang.Object

jsat.datatransform.featureselection.LRS

All Implemented Interfaces:

Serializable, Cloneable, DataTransform

public class LRS
extends Object
implements DataTransform

plus-L minus-R Selection (LRS) is a greedy method of selecting a subset of features to use for prediction. Its behavior is dependent upon whether L or R is the larger value. No mater what, L features will be greedily added to the set to decrease the error rate, and R features will be greedily removed while trying to maintain the error rate.
If L > R, then L-R features will be selected, the L step running first followed by R performing pruning on the found set.
If L < R, then D-R+L features will be selected, where D is the original number of features. First R features will be removed, and then L of the removed features will be added back to the final set.
L = R is not allowed.

Author:

Edward Raff

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
LRS(int L, int R, ClassificationDataSet cds, Classifier evaluater, int folds) Performs LRS feature selection for a classification problem
LRS(int L, int R, Classifier evaluater, int folds) Creates a LRS feature selection object for a classification problem
LRS(int L, int R, RegressionDataSet rds, Regressor evaluater, int folds) Performs LRS feature selection for a regression problem
LRS(int L, int R, Regressor evaluater, int folds) Creates a LRS feature selection object for a regression problem

Method Summary

Modifier and Type	Method and Description
LRS	clone()
void	fit(DataSet data) Fits this transform to the given dataset.
int	getFeaturesToAdd() Returns the number of features to add
int	getFeaturesToRemove() Returns the number of features to remove
int	getFolds()
Set<Integer>	getSelectedCategorical() Returns a copy of the set of categorical features selected by the search algorithm
Set<Integer>	getSelectedNumerical() Returns a copy of the set of numerical features selected by the search algorithm.
void	setFeaturesToAdd(int featuresToAdd) Sets the number of features to add (the L parameter).
void	setFeaturesToRemove(int featuresToRemove) Sets the number of features to remove (the R parameter).
void	setFolds(int folds) Sets the number of folds to use for cross validation when estimating the error rate
DataPoint	transform(DataPoint dp) Returns a new data point that is a transformation of the original data point.

Methods inherited from class java.lang.Object

equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

LRS

public LRS(int L,
           int R,
           Classifier evaluater,
           int folds)

Creates a LRS feature selection object for a classification problem

Parameters:

L - the number of features to greedily add

R - the number of features to greedily remove

evaluater - the classifier to use in determining accuracy given a feature subset

folds - the number of cross validation folds to use in selection

LRS

public LRS(int L,
           int R,
           ClassificationDataSet cds,
           Classifier evaluater,
           int folds)

Performs LRS feature selection for a classification problem

Parameters:

L - the number of features to greedily add

R - the number of features to greedily remove

cds - the data set to perform feature selection on

evaluater - the classifier to use in determining accuracy given a feature subset

folds - the number of cross validation folds to use in selection

LRS

public LRS(int L,
           int R,
           Regressor evaluater,
           int folds)

Creates a LRS feature selection object for a regression problem

Parameters:

L - the number of features to greedily add

R - the number of features to greedily remove

evaluater - the regressor to use in determining accuracy given a feature subset

folds - the number of cross validation folds to use in selection

LRS

public LRS(int L,
           int R,
           RegressionDataSet rds,
           Regressor evaluater,
           int folds)

Performs LRS feature selection for a regression problem

Parameters:

L - the number of features to greedily add

R - the number of features to greedily remove

rds - the data set to perform feature selection on

evaluater - the regressor to use in determining accuracy given a feature subset

folds - the number of cross validation folds to use in selection

Method Detail

transform

public DataPoint transform(DataPoint dp)

Description copied from interface: DataTransform

Returns a new data point that is a transformation of the original data point. This new data point is a different object, but may contain the same references as the original data point. It is not guaranteed that you can mutate the transformed point without having a side effect on the original point.

Specified by:

transform in interface DataTransform

Parameters:

dp - the data point to apply a transformation to

Returns:

a transformed data point

clone

public LRS clone()

Specified by:

clone in interface DataTransform

Overrides:

clone in class Object

getSelectedCategorical

public Set<Integer> getSelectedCategorical()

Returns a copy of the set of categorical features selected by the search algorithm

Returns:

the set of categorical features to use

getSelectedNumerical

public Set<Integer> getSelectedNumerical()

Returns a copy of the set of numerical features selected by the search algorithm.

Returns:

the set of numeric features to use

fit

public void fit(DataSet data)

Description copied from interface: DataTransform

Fits this transform to the given dataset. Some transforms can only be learned from classification or regression datasets. If an incompatible dataset type is given, a FailedToFitException exception may be thrown.

Specified by:

fit in interface DataTransform

Parameters:

data - the dataset to fir this transform to

setFeaturesToAdd

public void setFeaturesToAdd(int featuresToAdd)

Sets the number of features to add (the L parameter).

Parameters:

featuresToAdd - the number of features to greedily add

getFeaturesToAdd

public int getFeaturesToAdd()

Returns the number of features to add

Returns:

the number of features to add

setFeaturesToRemove

public void setFeaturesToRemove(int featuresToRemove)

Sets the number of features to remove (the R parameter).

Parameters:

featuresToRemove - the number of features to greedily remove

getFeaturesToRemove

public int getFeaturesToRemove()

Returns the number of features to remove

Returns:

the number of features to remove

setFolds

public void setFolds(int folds)

Sets the number of folds to use for cross validation when estimating the error rate

Parameters:

folds - the number of folds to use for cross validation when estimating the error rate

getFolds

public int getFolds()

Returns:

the number of folds to use for cross validation when estimating the error rate