Serialized Form


Package weka.attributeSelection

Class weka.attributeSelection.ASEvaluation extends java.lang.Object implements Serializable

serialVersionUID: 2091705669885950849L

Class weka.attributeSelection.ASSearch extends java.lang.Object implements Serializable

serialVersionUID: 7591673350342236548L

Class weka.attributeSelection.AttributeSelection extends java.lang.Object implements Serializable

serialVersionUID: 4170171824147584330L

Serialized Fields

m_trainInstances

weka.core.Instances m_trainInstances
the instances to select attributes from


m_ASEvaluator

weka.attributeSelection.ASEvaluation m_ASEvaluator
the attribute/subset evaluator


m_searchMethod

weka.attributeSelection.ASSearch m_searchMethod
the search method


m_numFolds

int m_numFolds
the number of folds to use for cross validation


m_selectionResults

java.lang.StringBuffer m_selectionResults
holds a string describing the results of the attribute selection


m_doRank

boolean m_doRank
rank features (if allowed by the search method)


m_doXval

boolean m_doXval
do cross validation


m_seed

int m_seed
seed used to randomly shuffle instances for cross validation


m_numToSelect

int m_numToSelect
number of attributes requested from ranked results


m_selectedAttributeSet

int[] m_selectedAttributeSet
the selected attributes


m_attributeRanking

double[][] m_attributeRanking
the attribute indexes and associated merits if a ranking is produced


m_transformer

weka.attributeSelection.AttributeTransformer m_transformer
if a feature selection run involves an attribute transformer


m_attributeFilter

weka.filters.unsupervised.attribute.Remove m_attributeFilter
the attribute filter for processing instances with respect to the most recent feature selection run


m_rankResults

double[][] m_rankResults
hold statistics for repeated feature selection, such as under cross validation


m_subsetResults

double[] m_subsetResults

m_trials

int m_trials

Class weka.attributeSelection.AttributeSetEvaluator extends weka.attributeSelection.ASEvaluation implements Serializable

serialVersionUID: -5744881009422257389L

Class weka.attributeSelection.BestFirst extends weka.attributeSelection.ASSearch implements Serializable

serialVersionUID: 7841338689536821867L

Serialized Fields

m_maxStale

int m_maxStale
maximum number of stale nodes before terminating search


m_searchDirection

int m_searchDirection
0 == backward search, 1 == forward search, 2 == bidirectional


m_starting

int[] m_starting
holds an array of starting attributes


m_startRange

weka.core.Range m_startRange
holds the start set for the search as a Range


m_hasClass

boolean m_hasClass
does the data have a class


m_classIndex

int m_classIndex
holds the class index


m_numAttribs

int m_numAttribs
number of attributes in the data


m_totalEvals

int m_totalEvals
total number of subsets evaluated during a search


m_debug

boolean m_debug
for debugging


m_bestMerit

double m_bestMerit
holds the merit of the best subset found


m_cacheSize

int m_cacheSize
holds the maximum size of the lookup cache for evaluated subsets

Class weka.attributeSelection.BestFirst.Link2 extends java.lang.Object implements Serializable

serialVersionUID: -8236598311516351420L

Serialized Fields

m_data

java.lang.Object[] m_data

m_merit

double m_merit

Class weka.attributeSelection.BestFirst.LinkedList2 extends weka.core.FastVector implements Serializable

serialVersionUID: 3250538292330398929L

Serialized Fields

m_MaxSize

int m_MaxSize
Max number of elements in the list

Class weka.attributeSelection.CfsSubsetEval extends weka.attributeSelection.ASEvaluation implements Serializable

serialVersionUID: 747878400813276317L

Serialized Fields

m_trainInstances

weka.core.Instances m_trainInstances
The training instances


m_disTransform

weka.filters.supervised.attribute.Discretize m_disTransform
Discretise attributes when class in nominal


m_classIndex

int m_classIndex
The class index


m_isNumeric

boolean m_isNumeric
Is the class numeric


m_numAttribs

int m_numAttribs
Number of attributes in the training data


m_numInstances

int m_numInstances
Number of instances in the training data


m_missingSeparate

boolean m_missingSeparate
Treat missing values as separate values


m_locallyPredictive

boolean m_locallyPredictive
Include locally predictive attributes


m_corr_matrix

float[][] m_corr_matrix
Holds the matrix of attribute correlations


m_std_devs

double[] m_std_devs
Standard deviations of attributes (when using pearsons correlation)


m_c_Threshold

double m_c_Threshold
Threshold for admitting locally predictive features

Class weka.attributeSelection.GainRatioAttributeEval extends weka.attributeSelection.ASEvaluation implements Serializable

serialVersionUID: -8504656625598579926L

Serialized Fields

m_trainInstances

weka.core.Instances m_trainInstances
The training instances


m_classIndex

int m_classIndex
The class index


m_numAttribs

int m_numAttribs
The number of attributes


m_numInstances

int m_numInstances
The number of instances


m_numClasses

int m_numClasses
The number of classes


m_missing_merge

boolean m_missing_merge
Merge missing values

Class weka.attributeSelection.GreedyStepwise extends weka.attributeSelection.ASSearch implements Serializable

serialVersionUID: -6312951970168325471L

Serialized Fields

m_hasClass

boolean m_hasClass
does the data have a class


m_classIndex

int m_classIndex
holds the class index


m_numAttribs

int m_numAttribs
number of attributes in the data


m_rankingRequested

boolean m_rankingRequested
true if the user has requested a ranked list of attributes


m_doRank

boolean m_doRank
go from one side of the search space to the other in order to generate a ranking


m_doneRanking

boolean m_doneRanking
used to indicate whether or not ranking has been performed


m_threshold

double m_threshold
A threshold by which to discard attributes---used by the AttributeSelection module


m_numToSelect

int m_numToSelect
The number of attributes to select. -1 indicates that all attributes are to be retained. Has precedence over m_threshold


m_calculatedNumToSelect

int m_calculatedNumToSelect

m_bestMerit

double m_bestMerit
the merit of the best subset found


m_rankedAtts

double[][] m_rankedAtts
a ranked list of attribute indexes


m_rankedSoFar

int m_rankedSoFar

m_best_group

java.util.BitSet m_best_group
the best subset found


m_ASEval

weka.attributeSelection.ASEvaluation m_ASEval

m_Instances

weka.core.Instances m_Instances

m_startRange

weka.core.Range m_startRange
holds the start set for the search as a Range


m_starting

int[] m_starting
holds an array of starting attributes


m_backward

boolean m_backward
Use a backwards search instead of a forwards one


m_conservativeSelection

boolean m_conservativeSelection
If set then attributes will continue to be added during a forward search as long as the merit does not degrade

Class weka.attributeSelection.HoldOutSubsetEvaluator extends weka.attributeSelection.ASEvaluation implements Serializable

serialVersionUID: 8280529785412054174L

Class weka.attributeSelection.InfoGainAttributeEval extends weka.attributeSelection.ASEvaluation implements Serializable

serialVersionUID: -1949849512589218930L

Serialized Fields

m_missing_merge

boolean m_missing_merge
Treat missing values as a seperate value


m_Binarize

boolean m_Binarize
Just binarize numeric attributes


m_InfoGains

double[] m_InfoGains
The info gain for each attribute

Class weka.attributeSelection.LatentSemanticAnalysis extends weka.attributeSelection.UnsupervisedAttributeEvaluator implements Serializable

serialVersionUID: 176853600006522257L

Serialized Fields

m_trainInstances

weka.core.Instances m_trainInstances
The data to transform analyse/transform


m_trainHeader

weka.core.Instances m_trainHeader
Keep a copy for the class attribute (if set) and for checking for header compatibility


m_transformedFormat

weka.core.Instances m_transformedFormat
The header for the transformed data format


m_hasClass

boolean m_hasClass
Data has a class set


m_classIndex

int m_classIndex
Class index


m_numAttributes

int m_numAttributes
Number of attributes


m_numInstances

int m_numInstances
Number of instances


m_transpose

boolean m_transpose
Is transpose necessary because numAttributes < numInstances?


m_u

weka.core.matrix.Matrix m_u
Will hold the left singular vectors


m_s

weka.core.matrix.Matrix m_s
Will hold the singular values


m_v

weka.core.matrix.Matrix m_v
Will hold the right singular vectors


m_transformationMatrix

weka.core.matrix.Matrix m_transformationMatrix
Will hold the matrix used to transform instances to the new feature space


m_replaceMissingFilter

weka.filters.unsupervised.attribute.ReplaceMissingValues m_replaceMissingFilter
Filters for original data


m_normalizeFilter

weka.filters.unsupervised.attribute.Normalize m_normalizeFilter

m_nominalToBinaryFilter

weka.filters.unsupervised.attribute.NominalToBinary m_nominalToBinaryFilter

m_attributeFilter

weka.filters.unsupervised.attribute.Remove m_attributeFilter

m_outputNumAttributes

int m_outputNumAttributes
The number of attributes in the LSA transformed data


m_normalize

boolean m_normalize
Normalize the input data?


m_rank

double m_rank
The approximation rank to use (between 0 and 1 means coverage proportion)


m_sumSquaredSingularValues

double m_sumSquaredSingularValues
The sum of the squares of the singular values


m_actualRank

int m_actualRank
The actual rank number to use for computation


m_maxAttributesInName

int m_maxAttributesInName
Maximum number of attributes in the transformed attribute name

Class weka.attributeSelection.OneRAttributeEval extends weka.attributeSelection.ASEvaluation implements Serializable

serialVersionUID: 4386514823886856980L

Serialized Fields

m_trainInstances

weka.core.Instances m_trainInstances
The training instances


m_classIndex

int m_classIndex
The class index


m_numAttribs

int m_numAttribs
The number of attributes


m_numInstances

int m_numInstances
The number of instances


m_randomSeed

int m_randomSeed
Random number seed


m_folds

int m_folds
Number of folds for cross validation


m_evalUsingTrainingData

boolean m_evalUsingTrainingData
Use training data to evaluate merit rather than x-val


m_minBucketSize

int m_minBucketSize
Passed on to OneR

Class weka.attributeSelection.PrincipalComponents extends weka.attributeSelection.UnsupervisedAttributeEvaluator implements Serializable

serialVersionUID: -3675307197777734007L

Serialized Fields

m_trainInstances

weka.core.Instances m_trainInstances
The data to transform analyse/transform


m_trainHeader

weka.core.Instances m_trainHeader
Keep a copy for the class attribute (if set)


m_transformedFormat

weka.core.Instances m_transformedFormat
The header for the transformed data format


m_originalSpaceFormat

weka.core.Instances m_originalSpaceFormat
The header for data transformed back to the original space


m_hasClass

boolean m_hasClass
Data has a class set


m_classIndex

int m_classIndex
Class index


m_numAttribs

int m_numAttribs
Number of attributes


m_numInstances

int m_numInstances
Number of instances


m_correlation

double[][] m_correlation
Correlation/covariance matrix for the original data


m_means

double[] m_means

m_stdDevs

double[] m_stdDevs

m_center

boolean m_center
If true, center (rather than standardize) the data and compute PCA from covariance (rather than correlation) matrix.


m_eigenvectors

double[][] m_eigenvectors
Will hold the unordered linear transformations of the (normalized) original data


m_eigenvalues

double[] m_eigenvalues
Eigenvalues for the corresponding eigenvectors


m_sortedEigens

int[] m_sortedEigens
Sorted eigenvalues


m_sumOfEigenValues

double m_sumOfEigenValues
sum of the eigenvalues


m_replaceMissingFilter

weka.filters.unsupervised.attribute.ReplaceMissingValues m_replaceMissingFilter
Filters for original data


m_nominalToBinFilter

weka.filters.unsupervised.attribute.NominalToBinary m_nominalToBinFilter

m_attributeFilter

weka.filters.unsupervised.attribute.Remove m_attributeFilter

m_centerFilter

weka.filters.unsupervised.attribute.Center m_centerFilter

m_standardizeFilter

weka.filters.unsupervised.attribute.Standardize m_standardizeFilter

m_outputNumAtts

int m_outputNumAtts
The number of attributes in the pc transformed data


m_coverVariance

double m_coverVariance
the amount of variance to cover in the original data when retaining the best n PC's


m_transBackToOriginal

boolean m_transBackToOriginal
transform the data through the pc space and back to the original space ?


m_maxAttrsInName

int m_maxAttrsInName
maximum number of attributes in the transformed attribute name


m_eTranspose

double[][] m_eTranspose
holds the transposed eigenvectors for converting back to the original space

Class weka.attributeSelection.Ranker extends weka.attributeSelection.ASSearch implements Serializable

serialVersionUID: -9086714848510751934L

Serialized Fields

m_starting

int[] m_starting
Holds the starting set as an array of attributes


m_startRange

weka.core.Range m_startRange
Holds the start set for the search as a range


m_attributeList

int[] m_attributeList
Holds the ordered list of attributes


m_attributeMerit

double[] m_attributeMerit
Holds the list of attribute merit scores


m_hasClass

boolean m_hasClass
Data has class attribute---if unsupervised evaluator then no class


m_classIndex

int m_classIndex
Class index of the data if supervised evaluator


m_numAttribs

int m_numAttribs
The number of attribtes


m_threshold

double m_threshold
A threshold by which to discard attributes---used by the AttributeSelection module


m_numToSelect

int m_numToSelect
The number of attributes to select. -1 indicates that all attributes are to be retained. Has precedence over m_threshold


m_calculatedNumToSelect

int m_calculatedNumToSelect
Used to compute the number to select

Class weka.attributeSelection.ReliefFAttributeEval extends weka.attributeSelection.ASEvaluation implements Serializable

serialVersionUID: -8422186665795839379L

Serialized Fields

m_trainInstances

weka.core.Instances m_trainInstances
The training instances


m_classIndex

int m_classIndex
The class index


m_numAttribs

int m_numAttribs
The number of attributes


m_numInstances

int m_numInstances
The number of instances


m_numericClass

boolean m_numericClass
Numeric class


m_numClasses

int m_numClasses
The number of classes if class is nominal


m_ndc

double m_ndc
Used to hold the probability of a different class val given nearest instances (numeric class)


m_nda

double[] m_nda
Used to hold the prob of different value of an attribute given nearest instances (numeric class case)


m_ndcda

double[] m_ndcda
Used to hold the prob of a different class val and different att val given nearest instances (numeric class case)


m_weights

double[] m_weights
Holds the weights that relief assigns to attributes


m_classProbs

double[] m_classProbs
Prior class probabilities (discrete class case)


m_sampleM

int m_sampleM
The number of instances to sample when estimating attributes default == -1, use all instances


m_Knn

int m_Knn
The number of nearest hits/misses


m_karray

double[][][] m_karray
k nearest scores + instance indexes for n classes


m_maxArray

double[] m_maxArray
Upper bound for numeric attributes


m_minArray

double[] m_minArray
Lower bound for numeric attributes


m_worst

double[] m_worst
Keep track of the farthest instance for each class


m_index

int[] m_index
Index in the m_karray of the farthest instance for each class


m_stored

int[] m_stored
Number of nearest neighbours stored of each class


m_seed

int m_seed
Random number seed used for sampling instances


m_weightsByRank

double[] m_weightsByRank
used to (optionally) weight nearest neighbours by their distance from the instance in question. Each entry holds exp(-((rank(r_i, i_j)/sigma)^2)) where rank(r_i,i_j) is the rank of instance i_j in a sequence of instances ordered by the distance from r_i. sigma is a user defined parameter, default=20


m_sigma

int m_sigma

m_weightByDistance

boolean m_weightByDistance
Weight by distance rather than equal weights

Class weka.attributeSelection.SymmetricalUncertAttributeEval extends weka.attributeSelection.ASEvaluation implements Serializable

serialVersionUID: -8096505776132296416L

Serialized Fields

m_trainInstances

weka.core.Instances m_trainInstances
The training instances


m_classIndex

int m_classIndex
The class index


m_numAttribs

int m_numAttribs
The number of attributes


m_numInstances

int m_numInstances
The number of instances


m_numClasses

int m_numClasses
The number of classes


m_missing_merge

boolean m_missing_merge
Treat missing values as a seperate value

Class weka.attributeSelection.UnsupervisedAttributeEvaluator extends weka.attributeSelection.ASEvaluation implements Serializable

serialVersionUID: -4100897318675336178L

Class weka.attributeSelection.UnsupervisedSubsetEvaluator extends weka.attributeSelection.ASEvaluation implements Serializable

serialVersionUID: 627934376267488763L

Class weka.attributeSelection.WrapperSubsetEval extends weka.attributeSelection.ASEvaluation implements Serializable

serialVersionUID: -4573057658746728675L

Serialized Fields

m_trainInstances

weka.core.Instances m_trainInstances
training instances


m_classIndex

int m_classIndex
class index


m_numAttribs

int m_numAttribs
number of attributes in the training data


m_numInstances

int m_numInstances
number of instances in the training data


m_Evaluation

weka.classifiers.Evaluation m_Evaluation
holds an evaluation object


m_BaseClassifier

weka.classifiers.Classifier m_BaseClassifier
holds the base classifier object


m_folds

int m_folds
number of folds to use for cross validation


m_seed

int m_seed
random number seed


m_threshold

double m_threshold
the threshold by which to do further cross validations when estimating the accuracy of a subset


m_evaluationMeasure

int m_evaluationMeasure
The evaluation measure to use