def featureSelection(self):
alg_search = ASSearch(
classname="weka.attributeSelection.GeneticSearch",
options=["-Z", "1024", "-G", "20", "-C", "0.6", "-M", "0.3"])
alg_evaluation = ASEvaluation(
classname="weka.attributeSelection.CfsSubsetEval",
options=["-P", "1", "-E", "1"])
feature_selection = AttributeSelection()
feature_selection.search(alg_search)
feature_selection.evaluator(alg_evaluation)
feature_selection.select_attributes(self.original_data)
self.selected_features = feature_selection.selected_attributes
self.num_features = feature_selection.number_attributes_selected
self.data_selected = feature_selection.reduce_dimensionality(
self.original_data)
cls = Classifier(classname=classifier)
evl = Evaluation(data)
evl.crossvalidate_model(cls, data, 10, Random(1))
print("%s: %0.0f%%" % (classifier, evl.percent_correct()))
# wrapper
for classifier in classifiers:
aseval = ASEvaluation(classname="weka.attributeSelection.WrapperSubsetEval",
options=["-B", classifier])
assearch = ASSearch(classname="weka.attributeSelection.BestFirst",
options=[])
attsel = AttributeSelection()
attsel.set_evaluator(aseval)
attsel.set_search(assearch)
attsel.select_attributes(data)
reduced = attsel.reduce_dimensionality(data)
cls = Classifier(classname=classifier)
evl = Evaluation(reduced)
evl.crossvalidate_model(cls, reduced, 10, Random(1))
print("%s (reduced): %0.0f%%" % (classifier, evl.percent_correct()))
# meta-classifier
for wrappercls in classifiers:
for basecls in classifiers:
meta = SingleClassifierEnhancer(classname="weka.classifiers.meta.AttributeSelectedClassifier")
meta.set_options(
["-E", "weka.attributeSelection.WrapperSubsetEval -B " + wrappercls,
"-S", "weka.attributeSelection.BestFirst",
"-W", basecls])
evl = Evaluation(data)
evl = Evaluation(data)
evl.crossvalidate_model(cls, data, 10, Random(1))
print("%s: %0.0f%%" % (classifier, evl.percent_correct))
# wrapper
for classifier in classifiers:
aseval = ASEvaluation(
classname="weka.attributeSelection.WrapperSubsetEval",
options=["-B", classifier])
assearch = ASSearch(classname="weka.attributeSelection.BestFirst",
options=[])
attsel = AttributeSelection()
attsel.evaluator(aseval)
attsel.search(assearch)
attsel.select_attributes(data)
reduced = attsel.reduce_dimensionality(data)
cls = Classifier(classname=classifier)
evl = Evaluation(reduced)
evl.crossvalidate_model(cls, reduced, 10, Random(1))
print("%s (reduced): %0.0f%%" % (classifier, evl.percent_correct))
# meta-classifier
for wrappercls in classifiers:
for basecls in classifiers:
meta = SingleClassifierEnhancer(
classname="weka.classifiers.meta.AttributeSelectedClassifier")
meta.options = \
["-E", "weka.attributeSelection.WrapperSubsetEval -B " + wrappercls,
"-S", "weka.attributeSelection.BestFirst",
"-W", basecls]
