注意
点击此处下载完整示例代码或通过 Binder 在浏览器中运行此示例
随机搜索¶
auto-sklearn 的一个关键特性是自动通过 SMAC 优化超参数，SMAC 在此处介绍。此外，也可以使用随机搜索代替 SMAC，如下面的示例所示。此外，本示例还演示了如何使用随机在线激进竞速 (ROAR) 作为另一种替代优化策略。这两个示例旨在展示 auto-sklearn 中的优化策略如何进行调整。
from pprint import pprint
import sklearn.model_selection
import sklearn.datasets
import sklearn.metrics
from smac.facade.roar_facade import ROAR
from smac.scenario.scenario import Scenario
import autosklearn.classification
数据加载¶
X, y = sklearn.datasets.load_breast_cancer(return_X_y=True)
X_train, X_test, y_train, y_test = sklearn.model_selection.train_test_split(
X, y, random_state=1
)
使用 ROAR 拟合分类器¶
def get_roar_object_callback(
scenario_dict,
seed,
ta,
ta_kwargs,
metalearning_configurations,
n_jobs,
dask_client,
multi_objective_algorithm, # This argument will be ignored as ROAR does not yet support multi-objective optimization
multi_objective_kwargs,
):
"""Random online adaptive racing."""
if n_jobs > 1 or (dask_client and len(dask_client.nthreads()) > 1):
raise ValueError(
"Please make sure to guard the code invoking Auto-sklearn by "
"`if __name__ == '__main__'` and remove this exception."
)
scenario = Scenario(scenario_dict)
return ROAR(
scenario=scenario,
rng=seed,
tae_runner=ta,
tae_runner_kwargs=ta_kwargs,
run_id=seed,
dask_client=dask_client,
n_jobs=n_jobs,
)
automl = autosklearn.classification.AutoSklearnClassifier(
time_left_for_this_task=60,
per_run_time_limit=15,
tmp_folder="/tmp/autosklearn_random_search_example_tmp",
initial_configurations_via_metalearning=0,
# The callback to get the SMAC object
get_smac_object_callback=get_roar_object_callback,
)
automl.fit(X_train, y_train, dataset_name="breast_cancer")
print("#" * 80)
print("Results for ROAR.")
# Print the final ensemble constructed by auto-sklearn via ROAR.
pprint(automl.show_models(), indent=4)
predictions = automl.predict(X_test)
# Print statistics about the auto-sklearn run such as number of
# iterations, number of models failed with a time out.
print(automl.sprint_statistics())
print("Accuracy score", sklearn.metrics.accuracy_score(y_test, predictions))
################################################################################
Results for ROAR.
{ 2: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d3b2c4c0>,
'cost': 0.028368794326241176,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d1e35a30>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d3b2c280>,
'model_id': 2,
'rank': 1,
'sklearn_classifier': RandomForestClassifier(max_features=5, n_estimators=512, n_jobs=1,
random_state=1, warm_start=True)},
3: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1bb37c0>,
'cost': 0.07092198581560283,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d4452910>,
'ensemble_weight': 0.04,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d4495520>,
'model_id': 3,
'rank': 2,
'sklearn_classifier': PassiveAggressiveClassifier(C=0.10318256510142626, average=True, max_iter=32,
random_state=1, tol=0.0013607858153657413,
warm_start=True)},
4: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1b2b9a0>,
'cost': 0.08510638297872342,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d3b2c8e0>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d154f970>,
'model_id': 4,
'rank': 3,
'sklearn_classifier': PassiveAggressiveClassifier(C=0.000505280701827856, max_iter=32, random_state=1,
tol=0.06421873693006744, warm_start=True)},
5: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1417490>,
'cost': 0.021276595744680882,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d1aa6e50>,
'ensemble_weight': 0.04,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1417100>,
'model_id': 5,
'rank': 4,
'sklearn_classifier': HistGradientBoostingClassifier(early_stopping=True,
l2_regularization=1.4469619760002166e-09,
learning_rate=0.12309321429220763, max_iter=256,
max_leaf_nodes=10, min_samples_leaf=17,
n_iter_no_change=16, random_state=1,
validation_fraction=None, warm_start=True)},
7: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d458d8e0>,
'cost': 0.36879432624113473,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d2413730>,
'ensemble_weight': 0.04,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d4547af0>,
'model_id': 7,
'rank': 5,
'sklearn_classifier': BernoulliNB(alpha=0.020898102999400082)},
8: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1a524c0>,
'cost': 0.028368794326241176,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d4133610>,
'ensemble_weight': 0.1,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1a52c70>,
'model_id': 8,
'rank': 6,
'sklearn_classifier': PassiveAggressiveClassifier(C=5.4951544593926646e-05, average=True, max_iter=16,
random_state=1, tol=0.012344344043531127,
warm_start=True)},
9: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1c152b0>,
'cost': 0.12056737588652477,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d6234430>,
'ensemble_weight': 0.08,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1c15940>,
'model_id': 9,
'rank': 7,
'sklearn_classifier': PassiveAggressiveClassifier(C=0.09581923415311244, max_iter=64, random_state=1,
tol=0.000145918046557035, warm_start=True)},
10: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d411bfd0>,
'cost': 0.16312056737588654,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d614df70>,
'ensemble_weight': 0.04,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d411b700>,
'model_id': 10,
'rank': 8,
'sklearn_classifier': ExtraTreesClassifier(criterion='entropy', max_features=2, min_samples_leaf=11,
min_samples_split=19, n_estimators=512, n_jobs=1,
random_state=1, warm_start=True)},
11: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05ca2e0b50>,
'cost': 0.028368794326241176,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d3d35e50>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d681f490>,
'model_id': 11,
'rank': 9,
'sklearn_classifier': LinearDiscriminantAnalysis(shrinkage='auto', solver='lsqr',
tol=0.039339489121391846)},
12: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d2403640>,
'cost': 0.13475177304964536,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d45bd820>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05cb155400>,
'model_id': 12,
'rank': 10,
'sklearn_classifier': LinearDiscriminantAnalysis(tol=0.0004182629077658861)},
13: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d61cdac0>,
'cost': 0.36879432624113473,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d25bfe20>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d3f33850>,
'model_id': 13,
'rank': 11,
'sklearn_classifier': ExtraTreesClassifier(bootstrap=True, criterion='entropy', max_features=1,
min_samples_leaf=20, min_samples_split=20,
n_estimators=512, n_jobs=1, random_state=1,
warm_start=True)},
14: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1ea3f70>,
'cost': 0.028368794326241176,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05cb2442b0>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1ea36a0>,
'model_id': 14,
'rank': 12,
'sklearn_classifier': LinearDiscriminantAnalysis(shrinkage='auto', solver='lsqr',
tol=0.013943862590174107)},
15: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d60168e0>,
'cost': 0.16312056737588654,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d248cbe0>,
'ensemble_weight': 0.04,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1a42040>,
'model_id': 15,
'rank': 13,
'sklearn_classifier': DecisionTreeClassifier(class_weight='balanced', criterion='entropy',
max_depth=1, min_samples_leaf=3, min_samples_split=13,
random_state=1)},
16: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1c00d90>,
'cost': 0.06382978723404253,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05c6e48fd0>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1c00340>,
'model_id': 16,
'rank': 14,
'sklearn_classifier': GaussianNB()},
17: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05ce652e20>,
'cost': 0.07801418439716312,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d43afc40>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05ce652d00>,
'model_id': 17,
'rank': 15,
'sklearn_classifier': PassiveAggressiveClassifier(C=0.04885675602027957, average=True, max_iter=64,
random_state=1, tol=0.0002678322187543955,
warm_start=True)},
18: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d3e68a90>,
'cost': 0.1063829787234043,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d21436d0>,
'ensemble_weight': 0.1,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d3e68460>,
'model_id': 18,
'rank': 16,
'sklearn_classifier': SGDClassifier(alpha=0.0002548862311968591, average=True,
eta0=4.278565071229579e-05, learning_rate='constant',
loss='squared_hinge', max_iter=1024, random_state=1,
tol=5.482673674397374e-05, warm_start=True)},
19: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d684e610>,
'cost': 0.09219858156028371,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d0f2ebe0>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d684eca0>,
'model_id': 19,
'rank': 17,
'sklearn_classifier': AdaBoostClassifier(base_estimator=DecisionTreeClassifier(max_depth=7),
learning_rate=0.035867722843024794, n_estimators=70,
random_state=1)},
23: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d19ad940>,
'cost': 0.07801418439716312,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05cc69a1c0>,
'ensemble_weight': 0.04,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d19ad2b0>,
'model_id': 23,
'rank': 18,
'sklearn_classifier': MLPClassifier(activation='tanh', alpha=0.02522490081077779, beta_1=0.999,
beta_2=0.9, early_stopping=True, hidden_layer_sizes=(35,),
learning_rate_init=0.026536043936270744, max_iter=32,
n_iter_no_change=32, random_state=1, verbose=0, warm_start=True)},
24: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1b78d90>,
'cost': 0.028368794326241176,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d2001670>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1b787c0>,
'model_id': 24,
'rank': 19,
'sklearn_classifier': LinearDiscriminantAnalysis(shrinkage='auto', solver='lsqr',
tol=0.001678686694320821)},
25: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d62b9e50>,
'cost': 0.021276595744680882,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d1efefd0>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d62b9400>,
'model_id': 25,
'rank': 20,
'sklearn_classifier': HistGradientBoostingClassifier(early_stopping=True,
l2_regularization=2.1928852428306468e-08,
learning_rate=0.29557366541255203, max_iter=64,
max_leaf_nodes=70, min_samples_leaf=24,
random_state=1,
validation_fraction=0.13252036572629863,
warm_start=True)},
26: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d61d3700>,
'cost': 0.08510638297872342,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d1df5100>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d154f040>,
'model_id': 26,
'rank': 21,
'sklearn_classifier': KNeighborsClassifier(n_neighbors=80, p=1)}}
auto-sklearn results:
Dataset name: breast_cancer
Metric: accuracy
Best validation score: 0.978723
Number of target algorithm runs: 26
Number of successful target algorithm runs: 25
Number of crashed target algorithm runs: 0
Number of target algorithms that exceeded the time limit: 1
Number of target algorithms that exceeded the memory limit: 0
Accuracy score 0.965034965034965
使用随机搜索拟合分类器¶
def get_random_search_object_callback(
scenario_dict,
seed,
ta,
ta_kwargs,
metalearning_configurations,
n_jobs,
dask_client,
multi_objective_algorithm, # This argument will be ignored as ROAR does not yet support multi-objective optimization
multi_objective_kwargs,
):
"""Random search"""
if n_jobs > 1 or (dask_client and len(dask_client.nthreads()) > 1):
raise ValueError(
"Please make sure to guard the code invoking Auto-sklearn by "
"`if __name__ == '__main__'` and remove this exception."
)
scenario_dict["minR"] = len(scenario_dict["instances"])
scenario_dict["initial_incumbent"] = "RANDOM"
scenario = Scenario(scenario_dict)
return ROAR(
scenario=scenario,
rng=seed,
tae_runner=ta,
tae_runner_kwargs=ta_kwargs,
run_id=seed,
dask_client=dask_client,
n_jobs=n_jobs,
)
automl = autosklearn.classification.AutoSklearnClassifier(
time_left_for_this_task=60,
per_run_time_limit=15,
tmp_folder="/tmp/autosklearn_random_search_example_tmp",
initial_configurations_via_metalearning=0,
# Passing the callback to get the SMAC object
get_smac_object_callback=get_random_search_object_callback,
)
automl.fit(X_train, y_train, dataset_name="breast_cancer")
print("#" * 80)
print("Results for random search.")
# Print the final ensemble constructed by auto-sklearn via random search.
pprint(automl.show_models(), indent=4)
# Print statistics about the auto-sklearn run such as number of
# iterations, number of models failed with a time out.
print(automl.sprint_statistics())
predictions = automl.predict(X_test)
print("Accuracy score", sklearn.metrics.accuracy_score(y_test, predictions))
################################################################################
Results for random search.
{ 3: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d618d640>,
'cost': 0.08510638297872342,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d618d820>,
'ensemble_weight': 0.04,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d618df40>,
'model_id': 3,
'rank': 1,
'sklearn_classifier': PassiveAggressiveClassifier(C=0.000505280701827856, max_iter=32, random_state=1,
tol=0.06421873693006744, warm_start=True)},
4: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1df5070>,
'cost': 0.021276595744680882,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d16ce850>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1df5ca0>,
'model_id': 4,
'rank': 2,
'sklearn_classifier': HistGradientBoostingClassifier(early_stopping=True,
l2_regularization=1.4469619760002166e-09,
learning_rate=0.12309321429220763, max_iter=256,
max_leaf_nodes=10, min_samples_leaf=17,
n_iter_no_change=16, random_state=1,
validation_fraction=None, warm_start=True)},
5: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d684ea90>,
'cost': 0.028368794326241176,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05cc052df0>,
'ensemble_weight': 0.14,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d684e070>,
'model_id': 5,
'rank': 3,
'sklearn_classifier': RandomForestClassifier(max_features=5, n_estimators=512, n_jobs=1,
random_state=1, warm_start=True)},
6: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d3e68190>,
'cost': 0.028368794326241176,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d611da60>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1d96220>,
'model_id': 6,
'rank': 4,
'sklearn_classifier': LinearDiscriminantAnalysis(tol=0.08622878828999171)},
7: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d4024820>,
'cost': 0.36879432624113473,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d4573b20>,
'ensemble_weight': 0.04,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d6403370>,
'model_id': 7,
'rank': 5,
'sklearn_classifier': BernoulliNB(alpha=0.020898102999400082)},
9: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d103b790>,
'cost': 0.12056737588652477,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d448d3d0>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d103bd30>,
'model_id': 9,
'rank': 6,
'sklearn_classifier': PassiveAggressiveClassifier(C=0.09581923415311244, max_iter=64, random_state=1,
tol=0.000145918046557035, warm_start=True)},
10: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1ad43d0>,
'cost': 0.16312056737588654,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05cb27baf0>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1ad4310>,
'model_id': 10,
'rank': 7,
'sklearn_classifier': ExtraTreesClassifier(criterion='entropy', max_features=2, min_samples_leaf=11,
min_samples_split=19, n_estimators=512, n_jobs=1,
random_state=1, warm_start=True)},
11: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d61e5be0>,
'cost': 0.028368794326241176,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d221fe50>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d457f9a0>,
'model_id': 11,
'rank': 8,
'sklearn_classifier': LinearDiscriminantAnalysis(shrinkage='auto', solver='lsqr',
tol=0.039339489121391846)},
12: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d25bf490>,
'cost': 0.13475177304964536,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d1ea3460>,
'ensemble_weight': 0.08,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1e27820>,
'model_id': 12,
'rank': 9,
'sklearn_classifier': LinearDiscriminantAnalysis(tol=0.0004182629077658861)},
13: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d0caa700>,
'cost': 0.36879432624113473,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d62fd250>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d0caae20>,
'model_id': 13,
'rank': 10,
'sklearn_classifier': ExtraTreesClassifier(bootstrap=True, criterion='entropy', max_features=1,
min_samples_leaf=20, min_samples_split=20,
n_estimators=512, n_jobs=1, random_state=1,
warm_start=True)},
14: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d0f45820>,
'cost': 0.028368794326241176,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05cc0528b0>,
'ensemble_weight': 0.04,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d45474c0>,
'model_id': 14,
'rank': 11,
'sklearn_classifier': LinearDiscriminantAnalysis(shrinkage='auto', solver='lsqr',
tol=0.013943862590174107)},
15: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05cc79a910>,
'cost': 0.16312056737588654,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d1ef6310>,
'ensemble_weight': 0.04,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d3d93940>,
'model_id': 15,
'rank': 12,
'sklearn_classifier': DecisionTreeClassifier(class_weight='balanced', criterion='entropy',
max_depth=1, min_samples_leaf=3, min_samples_split=13,
random_state=1)},
16: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1c0eeb0>,
'cost': 0.06382978723404253,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d2165070>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1c0e7c0>,
'model_id': 16,
'rank': 13,
'sklearn_classifier': GaussianNB()},
17: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1e355b0>,
'cost': 0.07801418439716312,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d0fc0790>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1e35850>,
'model_id': 17,
'rank': 14,
'sklearn_classifier': PassiveAggressiveClassifier(C=0.04885675602027957, average=True, max_iter=64,
random_state=1, tol=0.0002678322187543955,
warm_start=True)},
18: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d3dcb370>,
'cost': 0.1063829787234043,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05ca268730>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d3dcb190>,
'model_id': 18,
'rank': 15,
'sklearn_classifier': SGDClassifier(alpha=0.0002548862311968591, average=True,
eta0=4.278565071229579e-05, learning_rate='constant',
loss='squared_hinge', max_iter=1024, random_state=1,
tol=5.482673674397374e-05, warm_start=True)},
19: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05ca2e0700>,
'cost': 0.09219858156028371,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d0ccb430>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05ca2e0340>,
'model_id': 19,
'rank': 16,
'sklearn_classifier': AdaBoostClassifier(base_estimator=DecisionTreeClassifier(max_depth=7),
learning_rate=0.035867722843024794, n_estimators=70,
random_state=1)},
20: { 'balancing': Balancing(random_state=1),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d3b2c4f0>,
'cost': 0.07092198581560283,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d0fd70a0>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d3b2c970>,
'model_id': 20,
'rank': 17,
'sklearn_classifier': AdaBoostClassifier(algorithm='SAMME',
base_estimator=DecisionTreeClassifier(max_depth=9),
learning_rate=0.21684295401135145, n_estimators=72,
random_state=1)},
23: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1c62880>,
'cost': 0.07801418439716312,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d2413fd0>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1c62580>,
'model_id': 23,
'rank': 18,
'sklearn_classifier': MLPClassifier(activation='tanh', alpha=0.02522490081077779, beta_1=0.999,
beta_2=0.9, early_stopping=True, hidden_layer_sizes=(35,),
learning_rate_init=0.026536043936270744, max_iter=32,
n_iter_no_change=32, random_state=1, verbose=0, warm_start=True)},
24: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d624be50>,
'cost': 0.028368794326241176,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d3aee2e0>,
'ensemble_weight': 0.06,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d624bd90>,
'model_id': 24,
'rank': 19,
'sklearn_classifier': LinearDiscriminantAnalysis(shrinkage='auto', solver='lsqr',
tol=0.001678686694320821)},
25: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d1bf6280>,
'cost': 0.021276595744680882,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d0c32c40>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d1bf6d00>,
'model_id': 25,
'rank': 20,
'sklearn_classifier': HistGradientBoostingClassifier(early_stopping=True,
l2_regularization=2.1928852428306468e-08,
learning_rate=0.29557366541255203, max_iter=64,
max_leaf_nodes=70, min_samples_leaf=24,
random_state=1,
validation_fraction=0.13252036572629863,
warm_start=True)},
26: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d62efb20>,
'cost': 0.08510638297872342,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d1c62ca0>,
'ensemble_weight': 0.02,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d3b0e820>,
'model_id': 26,
'rank': 21,
'sklearn_classifier': KNeighborsClassifier(n_neighbors=80, p=1)},
27: { 'balancing': Balancing(random_state=1, strategy='weighting'),
'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x7f05d438e1f0>,
'cost': 0.375886524822695,
'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x7f05d1b2bbe0>,
'ensemble_weight': 0.04,
'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x7f05d482d310>,
'model_id': 27,
'rank': 22,
'sklearn_classifier': BernoulliNB(alpha=9.144520523203797)}}
auto-sklearn results:
Dataset name: breast_cancer
Metric: accuracy
Best validation score: 0.978723
Number of target algorithm runs: 26
Number of successful target algorithm runs: 26
Number of crashed target algorithm runs: 0
Number of target algorithms that exceeded the time limit: 0
Number of target algorithms that exceeded the memory limit: 0
Accuracy score 0.972027972027972
脚本总运行时间： ( 1 分 56.774 秒)
