def test_QueryMarginSampling(self):
# init the AlExperiment
al = ClassicActiveLearning(self.__X,
self.__y,
model=RandomForestClassifier(
max_depth=2, random_state=0),
stopping_criteria=UnlabeledSetEmpty())
# create a kfold experiment
al.kfold(n_splits=10)
# set the query strategy
strategy = QueryMarginSampling()
al.set_query_strategy(strategy=QueryMarginSampling())
# set the metric for experiment.
al.set_performance_metric('accuracy_score')
# by default,run in multi-thread.
al.execute()
# get the experiemnt result
stateIO = al.get_experiment_result()
# get a brief description of the experiment
al.plot_learning_curve(title='Alexperiment result %s' %
strategy.query_function_name)
def test_QueryInstanceRandom(self):
# init the AlExperiment
al = ClassicActiveLearning(self.__X,
self.__y,
model=RandomForestClassifier(
max_depth=2, random_state=0),
stopping_criteria=MaxIteration(value=10))
# create a kfold experiment
al.kfold(n_splits=10)
# set the query strategy
strategy = QueryInstanceRandom()
al.set_query_strategy(strategy=QueryInstanceRandom())
# set the metric for experiment.
al.set_performance_metric('accuracy_score')
# Execute the experiment
al.execute(verbose=False)
# get the experiemnt result
# stateIO = al.get_experiment_result()
# get a brief description of the experiment
al.plot_learning_curve(title='Alexperiment result %s' %
strategy.query_function_name)