|
|||||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | ||||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |
java.lang.Objectrelational.classifier.Acmclassifier
Main class. All the procedures are called from here.
Field Summary | |
double |
acc
|
double[] |
accuracy
|
int |
dataSet
|
int |
dataSetSize
|
double |
f1
|
edu.uci.ics.jung.graph.impl.SparseGraph |
g
|
Iterative |
iter
|
double |
macro_f1
|
double |
macro_precision
|
double |
macro_recall
|
int |
maxIterations
|
int |
maxNumNeighbors
|
double |
micro_f1
|
double |
micro_precision
|
double |
micro_recall
|
double |
minWeight
|
int |
neighb
|
int |
numMax
|
double |
precision
|
java.util.HashMap |
probabilities
|
double |
recall
|
java.util.HashMap |
result
|
weka.core.FastVector |
test
|
int |
testSize
|
weka.core.FastVector |
train
|
int |
trainSize
|
static java.lang.String |
type
|
static java.util.HashMap |
weights
|
java.util.HashMap |
weightsMap
|
static java.lang.Object[] |
weightsTrain
|
Constructor Summary | |
Acmclassifier(edu.uci.ics.jung.graph.impl.SparseGraph g,
boolean multiLabel,
java.util.HashMap labels,
java.lang.String type)
|
Method Summary | |
void |
combineModelsVoting(java.util.HashMap[] models,
int numFolds)
Combine models by using Voting |
void |
combineModelsVotingMulti(int numFolds,
double th,
java.util.HashMap[] models)
Combine models by using Voting |
double |
evaluate(int numRuns,
java.lang.String algorithm,
java.util.HashMap labels,
weka.core.FastVector[] traintest,
java.util.HashMap weightsMap)
Calls the classification algorithms and evaluates the results |
void |
evaluateMultiLabel(java.lang.String algorithm,
int numRuns,
java.util.HashMap labels,
weka.core.FastVector[] traintest,
java.util.HashMap weightsMap,
double th)
Calls the classification algorithms and evaluates the results (in the case of multilabel problem) |
void |
evaluateWithStacking(int numFolds,
java.util.HashMap[] models)
Combine models by using Stacking |
void |
evaluateWithStackingMulti(int numFolds,
java.util.HashMap[] models,
double th)
Combine models by using Stacking |
java.util.HashMap |
initializeWithLocalClassification(weka.core.FastVector test,
java.util.HashMap probs,
double th,
int Class)
Initializes the test instances with the results of a local classifier (nominal initialization)(for binary classification) |
java.util.HashMap |
initializeWithLocalProbs(weka.core.FastVector test,
java.util.HashMap probs)
Initializes the test instances with the results of a local classifier |
java.util.HashMap |
initializeWithLocalProbsBin(weka.core.FastVector test,
java.util.HashMap probs,
int Class)
Initializes the test instances with the results of a local classifier (binary classification) |
java.util.HashMap |
initializeWithPriorClass(weka.core.FastVector test,
double[] priors)
Initializes the test instances with the prior class calculated from the train set |
java.util.HashMap |
initializeWithPriorClass(weka.core.FastVector test,
double[] priors,
int Class)
Initializes the test instances with the prior probability of the train set |
java.util.HashMap |
initializeWithPriors(weka.core.FastVector test,
double[] priors)
Initializes the test instances with the prior probability of the train set |
java.util.HashMap |
initializeWithPriors(weka.core.Instances test,
double[] priors)
Initializes the test instances with the prior probability of the train set |
java.util.HashMap |
initializeWithZero(weka.core.FastVector test)
Initializes the test instances with zero |
java.util.HashMap |
initializeWithZero(weka.core.Instances test)
Initializes the test instances with zero |
java.util.HashMap |
initializeWithZeroMulti(weka.core.FastVector test)
Initializes the test instances with zero |
static void |
main(java.lang.String[] args)
|
static weka.core.Instances |
setupInstances(weka.core.FastVector items,
java.util.HashMap labels)
Creates new instances |
weka.core.Instances |
setupInstancesBin(weka.core.FastVector items,
int Class)
Creates instances for binary classification (one-vs.rest approach) |
weka.core.Instances |
setupInstancesBinStack(weka.core.FastVector items,
int klasse,
java.util.HashMap[] models)
Creates new instances for Stacking (binary classification) |
static weka.core.Instances |
setupInstancesCosineSimBin(weka.core.FastVector items,
int Class,
java.util.HashMap sim)
Creates instances containing the cosine similarity to neighbors (IndRVS algorithm) |
static weka.core.Instances |
setupInstancesCount(weka.core.FastVector items,
java.util.HashMap count)
Creates new instances containing the values for the aggregation function WeightedCount |
static weka.core.Instances |
setupInstancesCountBin(weka.core.FastVector items,
int Class,
java.util.HashMap count,
java.lang.String traintest)
Creates new instances containing the values for the aggregation function WeightedCount (binary classification) |
static weka.core.Instances |
setupInstancesStack(weka.core.FastVector items,
java.util.HashMap labels,
java.util.HashMap[] models)
Creates new instances for Stacking |
static weka.core.Instances |
setupInstancesWithNeighbors(weka.core.FastVector items,
java.util.HashMap labels,
java.util.HashMap id2label,
int numMaxNeighbors,
java.util.HashMap weightsMap,
java.util.HashMap nodes,
java.lang.String traintest)
Creates new instances for WeightedNaiveBayes |
weka.core.Instances |
setupInstancesWithNeighborsBin(weka.core.FastVector items,
java.util.HashMap id2labels,
int numMaxNeighbors,
int klasse,
java.util.HashMap weightsMap,
java.util.HashMap nodes,
java.lang.String traintest)
Creates new instances WeightedNaiveBayes (binary classification) |
weka.core.Instance |
setupInstTest(BipartiteNode item,
java.util.HashMap id2labels,
weka.core.Instance v,
int klasse,
int numMaxNeighbors,
java.util.HashMap weightsMap)
|
static weka.core.Instance |
setupInstTestHet(BipartiteNode item,
java.util.HashMap id2label,
weka.core.Instance v,
int numMaxNeighbors,
java.util.HashMap weightsMap)
|
static weka.core.Instance |
setupInstTestHom(Node item,
java.util.HashMap id2label,
weka.core.Instance v,
int numMaxNeighbors,
java.util.HashMap weightsMap)
|
weka.core.Instance |
setupInstTestHom(Node item,
java.util.HashMap id2labels,
weka.core.Instance v,
int klasse,
int numMaxNeighbors,
java.util.HashMap weightsMap)
|
weka.core.Instance |
setupInstTrain(BipartiteNode item,
java.util.HashMap id2labels,
weka.core.Instance v,
int klasse,
int numMaxNeighbors,
java.util.HashMap weightsMap)
|
static weka.core.Instance |
setupInstTrainHet(BipartiteNode item,
java.util.HashMap id2label,
weka.core.Instance v,
int numMaxNeighbors,
java.util.HashMap weightsMap)
|
static weka.core.Instance |
setupInstTrainHom(Node item,
java.util.HashMap id2label,
weka.core.Instance v,
int numMaxNeighbors,
java.util.HashMap weightsMap)
|
weka.core.Instance |
setupInstTrainHom(Node item,
java.util.HashMap id2labels,
weka.core.Instance v,
int klasse,
int numMaxNeighbors,
java.util.HashMap weightsMap)
|
static weka.core.SparseInstance |
setupSparseInstHet(BipartiteNode item,
weka.core.SparseInstance v)
|
static weka.core.SparseInstance |
setupSparseInstHom(Node item,
weka.core.SparseInstance v)
|
Methods inherited from class java.lang.Object |
equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
Field Detail |
public weka.core.FastVector train
public weka.core.FastVector test
public java.util.HashMap result
public static java.lang.Object[] weightsTrain
public edu.uci.ics.jung.graph.impl.SparseGraph g
public int neighb
public int dataSet
public static java.util.HashMap weights
public double minWeight
public java.util.HashMap probabilities
public int maxIterations
public int numMax
public java.util.HashMap weightsMap
public double micro_f1
public double macro_f1
public double micro_recall
public Iterative iter
public double macro_recall
public double micro_precision
public double macro_precision
public double[] accuracy
public int dataSetSize
public int testSize
public int trainSize
public double acc
public double f1
public double recall
public double precision
public static java.lang.String type
public int maxNumNeighbors
Constructor Detail |
public Acmclassifier(edu.uci.ics.jung.graph.impl.SparseGraph g, boolean multiLabel, java.util.HashMap labels, java.lang.String type) throws java.io.IOException, javax.xml.parsers.ParserConfigurationException, org.xml.sax.SAXException
Method Detail |
public weka.core.Instances setupInstancesBin(weka.core.FastVector items, int Class)
items
- - data instancesClass
- - the category
public weka.core.Instances setupInstancesBinStack(weka.core.FastVector items, int klasse, java.util.HashMap[] models)
items
- - data instancesmodels
- - the models to be combined
public static weka.core.Instances setupInstancesStack(weka.core.FastVector items, java.util.HashMap labels, java.util.HashMap[] models)
items
- - data instanceslabels
- - classes contained in the data setmodels
- - the models to be combined
public static weka.core.Instances setupInstancesCosineSimBin(weka.core.FastVector items, int Class, java.util.HashMap sim)
items
- - data instancesClass
- - the categorysim
- - calculated similarities
public static weka.core.Instances setupInstancesCount(weka.core.FastVector items, java.util.HashMap count)
items
- - data instancescount
- - the weighted counts (regarding the classes of the neighbors)
public static weka.core.Instances setupInstancesCountBin(weka.core.FastVector items, int Class, java.util.HashMap count, java.lang.String traintest)
items
- - dataClass
- - the categorycount
- - the weighted counts (regarding the classes of the neighbors)
public static weka.core.Instances setupInstances(weka.core.FastVector items, java.util.HashMap labels)
items
- - datalabels
- - the classes contained in the data set
public static weka.core.SparseInstance setupSparseInstHet(BipartiteNode item, weka.core.SparseInstance v)
public static weka.core.SparseInstance setupSparseInstHom(Node item, weka.core.SparseInstance v)
public static weka.core.Instances setupInstancesWithNeighbors(weka.core.FastVector items, java.util.HashMap labels, java.util.HashMap id2label, int numMaxNeighbors, java.util.HashMap weightsMap, java.util.HashMap nodes, java.lang.String traintest)
items
- - datalabels
- - the classe of the data setid2label
- - contains initialization for test setnumMaxNeighbors
- - the maximal number of neighbors to be consideredweightsMap
- - contains the weigths of the graphnodes
- - nodes of the graphtraintest
- - file path to the training and test data
public static weka.core.Instance setupInstTrainHet(BipartiteNode item, java.util.HashMap id2label, weka.core.Instance v, int numMaxNeighbors, java.util.HashMap weightsMap)
public static weka.core.Instance setupInstTrainHom(Node item, java.util.HashMap id2label, weka.core.Instance v, int numMaxNeighbors, java.util.HashMap weightsMap)
public static weka.core.Instance setupInstTestHet(BipartiteNode item, java.util.HashMap id2label, weka.core.Instance v, int numMaxNeighbors, java.util.HashMap weightsMap)
public static weka.core.Instance setupInstTestHom(Node item, java.util.HashMap id2label, weka.core.Instance v, int numMaxNeighbors, java.util.HashMap weightsMap)
public weka.core.Instances setupInstancesWithNeighborsBin(weka.core.FastVector items, java.util.HashMap id2labels, int numMaxNeighbors, int klasse, java.util.HashMap weightsMap, java.util.HashMap nodes, java.lang.String traintest)
items
- - datanumMaxNeighbors
- - the maximal number of neighbors to be consideredweightsMap
- - contains the weigths of the graphnodes
- - nodes of the graphtraintest
- - file path to the training and test data
public weka.core.Instance setupInstTrain(BipartiteNode item, java.util.HashMap id2labels, weka.core.Instance v, int klasse, int numMaxNeighbors, java.util.HashMap weightsMap)
public weka.core.Instance setupInstTrainHom(Node item, java.util.HashMap id2labels, weka.core.Instance v, int klasse, int numMaxNeighbors, java.util.HashMap weightsMap)
public weka.core.Instance setupInstTest(BipartiteNode item, java.util.HashMap id2labels, weka.core.Instance v, int klasse, int numMaxNeighbors, java.util.HashMap weightsMap)
public weka.core.Instance setupInstTestHom(Node item, java.util.HashMap id2labels, weka.core.Instance v, int klasse, int numMaxNeighbors, java.util.HashMap weightsMap)
public double evaluate(int numRuns, java.lang.String algorithm, java.util.HashMap labels, weka.core.FastVector[] traintest, java.util.HashMap weightsMap) throws java.lang.Exception
numRuns
- - number of folds for cross validationalgorithm
- - the algorithm to be performedlabels
- - contains categoriestraintest
- - contains train and test setsweightsMap
- - contains the weigths of the graph
java.lang.Exception
public void evaluateMultiLabel(java.lang.String algorithm, int numRuns, java.util.HashMap labels, weka.core.FastVector[] traintest, java.util.HashMap weightsMap, double th) throws java.lang.Exception
algorithm
- - the algorithm to be performednumRuns
- - number of folds for cross validationlabels
- - contains categoriestraintest
- - contains train and test setsweightsMap
- - contains the weigths of the graphth
- - threshold
java.lang.Exception
public java.util.HashMap initializeWithLocalProbsBin(weka.core.FastVector test, java.util.HashMap probs, int Class)
test
- - test setprobs
- - results (probabilities) achieved by local classifierClass
- - category
public java.util.HashMap initializeWithLocalProbs(weka.core.FastVector test, java.util.HashMap probs)
test
- - test setprobs
- - results (probabilities) achieved by local classifier
public java.util.HashMap initializeWithLocalClassification(weka.core.FastVector test, java.util.HashMap probs, double th, int Class)
test
- - test setprobs
- - results (probabilities) achieved by local classifierth
- - thresholdClass
- - the category
public java.util.HashMap initializeWithPriors(weka.core.Instances test, double[] priors)
test
- - test setpriors
- - prior probability calculated from the train set
public java.util.HashMap initializeWithPriors(weka.core.FastVector test, double[] priors)
test
- - test setpriors
- - prior probability calculated from the train set
public java.util.HashMap initializeWithPriorClass(weka.core.FastVector test, double[] priors)
test
- - test setpriors
- - prior probability calculated from the train set
public java.util.HashMap initializeWithPriorClass(weka.core.FastVector test, double[] priors, int Class)
test
- - test setpriors
- - prior probability calculated from the train setClass
- - the category
public java.util.HashMap initializeWithZero(weka.core.Instances test)
test
- - test set
public java.util.HashMap initializeWithZero(weka.core.FastVector test)
test
- - test set
public java.util.HashMap initializeWithZeroMulti(weka.core.FastVector test)
test
- - test set
public void combineModelsVoting(java.util.HashMap[] models, int numFolds)
models
- - results of the base classifiersnumFolds
- - number of folds in cross-validationpublic void combineModelsVotingMulti(int numFolds, double th, java.util.HashMap[] models)
numFolds
- - number of folds in cross-validationth
- - thresholdmodels
- - results of the base classifierspublic void evaluateWithStackingMulti(int numFolds, java.util.HashMap[] models, double th) throws java.lang.Exception
numFolds
- - number of folds in cross-validationmodels
- - results of the base classifiersth
- - threshold
java.lang.Exception
public void evaluateWithStacking(int numFolds, java.util.HashMap[] models) throws java.lang.Exception
numFolds
- - number of folds in cross-validationmodels
- - results of the base classifiers
java.lang.Exception
public static void main(java.lang.String[] args) throws java.lang.Exception
java.lang.Exception
|
|||||||||||
PREV CLASS NEXT CLASS | FRAMES NO FRAMES | ||||||||||
SUMMARY: NESTED | FIELD | CONSTR | METHOD | DETAIL: FIELD | CONSTR | METHOD |