diff --git a/Auto-SKLearn_AutoML/Classification.ipynb b/Auto-SKLearn_AutoML/Classification.ipynb
index 0585d146407718f0c1f47c6321d846e3922e5a29..14c5640829dd24121a4c03d5e481f86e4940e8c5 100644
--- a/Auto-SKLearn_AutoML/Classification.ipynb
+++ b/Auto-SKLearn_AutoML/Classification.ipynb
@@ -12,26 +12,29 @@
},
{
"cell_type": "code",
- "execution_count": 1,
+ "execution_count": 12,
"id": "c69433ce",
"metadata": {},
"outputs": [],
"source": [
+ "# imports\n",
"from pprint import pprint\n",
"\n",
"import sklearn.datasets\n",
"import sklearn.metrics\n",
+ "import pickle\n",
"\n",
"import autosklearn.classification"
]
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 2,
"id": "2b1e1930",
"metadata": {},
"outputs": [],
"source": [
+ "# split the dataset\n",
"X, y = sklearn.datasets.load_breast_cancer(return_X_y=True)\n",
"X_train, X_test, y_train, y_test = \\\n",
" sklearn.model_selection.train_test_split(X, y, random_state=1)"
@@ -39,11 +42,34 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 3,
"id": "15e5f821",
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stderr",
+ "output_type": "stream",
+ "text": [
+ "/mnt/home/buckl113/anaconda3/envs/auto_sklearn_env/lib/python3.9/site-packages/autosklearn/metalearning/metalearning/meta_base.py:68: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n",
+ " self.metafeatures = self.metafeatures.append(metafeatures)\n",
+ "/mnt/home/buckl113/anaconda3/envs/auto_sklearn_env/lib/python3.9/site-packages/autosklearn/metalearning/metalearning/meta_base.py:72: FutureWarning: The frame.append method is deprecated and will be removed from pandas in a future version. Use pandas.concat instead.\n",
+ " self.algorithm_runs[metric].append(runs)\n"
+ ]
+ },
+ {
+ "data": {
+ "text/plain": [
+ "AutoSklearnClassifier(per_run_time_limit=30, time_left_for_this_task=120,\n",
+ " tmp_folder='/tmp/autosklearn_classification_example_tmp')"
+ ]
+ },
+ "execution_count": 3,
+ "metadata": {},
+ "output_type": "execute_result"
+ }
+ ],
"source": [
+ "# Fit the classifier\n",
"automl = autosklearn.classification.AutoSklearnClassifier(\n",
" time_left_for_this_task=120,\n",
" per_run_time_limit=30,\n",
@@ -54,34 +80,406 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 4,
"id": "2d4e4d9f",
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ " rank ensemble_weight type cost duration\n",
+ "model_id \n",
+ "7 1 0.08 extra_trees 0.014184 2.037889\n",
+ "16 2 0.04 gradient_boosting 0.021277 1.420893\n",
+ "2 3 0.02 random_forest 0.028369 2.230037\n",
+ "3 4 0.02 mlp 0.028369 2.112077\n",
+ "26 5 0.04 extra_trees 0.028369 2.946061\n",
+ "6 6 0.06 mlp 0.028369 1.590191\n",
+ "22 7 0.04 gradient_boosting 0.028369 1.567744\n",
+ "10 8 0.06 random_forest 0.028369 2.733248\n",
+ "11 9 0.04 random_forest 0.028369 3.057115\n",
+ "19 10 0.02 extra_trees 0.028369 3.806663\n",
+ "13 11 0.04 gradient_boosting 0.028369 2.005229\n",
+ "14 12 0.04 mlp 0.028369 3.018905\n",
+ "18 13 0.02 random_forest 0.035461 2.909402\n",
+ "17 14 0.02 gradient_boosting 0.035461 2.358046\n",
+ "12 15 0.02 gradient_boosting 0.035461 2.129865\n",
+ "5 16 0.10 random_forest 0.035461 2.896027\n",
+ "9 17 0.02 extra_trees 0.042553 2.569632\n",
+ "25 18 0.02 adaboost 0.042553 2.708276\n",
+ "27 19 0.04 extra_trees 0.042553 2.473708\n",
+ "23 20 0.06 mlp 0.049645 3.621525\n",
+ "32 21 0.04 extra_trees 0.063830 2.219691\n",
+ "24 22 0.06 random_forest 0.070922 2.306794\n",
+ "20 23 0.02 passive_aggressive 0.078014 0.973551\n",
+ "30 24 0.02 adaboost 0.078014 1.630544\n",
+ "29 25 0.06 gaussian_nb 0.141844 1.153570\n"
+ ]
+ }
+ ],
"source": [
+ "# Different Models run by autosklearn\n",
"print(automl.leaderboard())"
]
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 5,
"id": "72e580e7",
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "{ 2: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3ac8f13d0>,\n",
+ " 'cost': 0.028368794326241176,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3ac688a90>,\n",
+ " 'ensemble_weight': 0.02,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3ac8f1310>,\n",
+ " 'model_id': 2,\n",
+ " 'rank': 3,\n",
+ " 'sklearn_classifier': RandomForestClassifier(max_features=5, n_estimators=512, n_jobs=1,\n",
+ " random_state=1, warm_start=True)},\n",
+ " 3: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3ac95b820>,\n",
+ " 'cost': 0.028368794326241176,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3ac6541f0>,\n",
+ " 'ensemble_weight': 0.02,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3ac95b6a0>,\n",
+ " 'model_id': 3,\n",
+ " 'rank': 4,\n",
+ " 'sklearn_classifier': MLPClassifier(activation='tanh', alpha=0.0001363185819149026, beta_1=0.999,\n",
+ " beta_2=0.9, early_stopping=True,\n",
+ " hidden_layer_sizes=(115, 115, 115),\n",
+ " learning_rate_init=0.00018009776276177523, max_iter=32,\n",
+ " n_iter_no_change=32, random_state=1, verbose=0, warm_start=True)},\n",
+ " 5: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b0500820>,\n",
+ " 'cost': 0.03546099290780147,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3affaeeb0>,\n",
+ " 'ensemble_weight': 0.1,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b034d700>,\n",
+ " 'model_id': 5,\n",
+ " 'rank': 13,\n",
+ " 'sklearn_classifier': RandomForestClassifier(criterion='entropy', max_features=3, min_samples_leaf=2,\n",
+ " n_estimators=512, n_jobs=1, random_state=1,\n",
+ " warm_start=True)},\n",
+ " 6: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3af75d4c0>,\n",
+ " 'cost': 0.028368794326241176,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3ac8477c0>,\n",
+ " 'ensemble_weight': 0.06,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3af75d3a0>,\n",
+ " 'model_id': 6,\n",
+ " 'rank': 5,\n",
+ " 'sklearn_classifier': MLPClassifier(alpha=0.0017940473175767063, beta_1=0.999, beta_2=0.9,\n",
+ " early_stopping=True, hidden_layer_sizes=(101, 101),\n",
+ " learning_rate_init=0.0004684917334431039, max_iter=32,\n",
+ " n_iter_no_change=32, random_state=1, verbose=0, warm_start=True)},\n",
+ " 7: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3ac829e50>,\n",
+ " 'cost': 0.014184397163120588,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3ac1c6c40>,\n",
+ " 'ensemble_weight': 0.08,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3ac829ee0>,\n",
+ " 'model_id': 7,\n",
+ " 'rank': 1,\n",
+ " 'sklearn_classifier': ExtraTreesClassifier(max_features=34, min_samples_leaf=3, min_samples_split=11,\n",
+ " n_estimators=512, n_jobs=1, random_state=1,\n",
+ " warm_start=True)},\n",
+ " 9: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3afaa6100>,\n",
+ " 'cost': 0.04255319148936165,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3b0717b80>,\n",
+ " 'ensemble_weight': 0.02,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b0809c10>,\n",
+ " 'model_id': 9,\n",
+ " 'rank': 17,\n",
+ " 'sklearn_classifier': ExtraTreesClassifier(max_features=9, min_samples_split=10, n_estimators=512,\n",
+ " n_jobs=1, random_state=1, warm_start=True)},\n",
+ " 10: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3af765670>,\n",
+ " 'cost': 0.028368794326241176,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3ac75f1c0>,\n",
+ " 'ensemble_weight': 0.06,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3af765550>,\n",
+ " 'model_id': 10,\n",
+ " 'rank': 6,\n",
+ " 'sklearn_classifier': RandomForestClassifier(criterion='entropy', max_features=4, min_samples_split=6,\n",
+ " n_estimators=512, n_jobs=1, random_state=1,\n",
+ " warm_start=True)},\n",
+ " 11: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3af85dac0>,\n",
+ " 'cost': 0.028368794326241176,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3ae0ae9a0>,\n",
+ " 'ensemble_weight': 0.04,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3af85d940>,\n",
+ " 'model_id': 11,\n",
+ " 'rank': 7,\n",
+ " 'sklearn_classifier': RandomForestClassifier(criterion='entropy', max_features=23, min_samples_leaf=7,\n",
+ " n_estimators=512, n_jobs=1, random_state=1,\n",
+ " warm_start=True)},\n",
+ " 12: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b04ebca0>,\n",
+ " 'cost': 0.03546099290780147,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3b0045ca0>,\n",
+ " 'ensemble_weight': 0.02,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b04eb820>,\n",
+ " 'model_id': 12,\n",
+ " 'rank': 14,\n",
+ " 'sklearn_classifier': HistGradientBoostingClassifier(early_stopping=True,\n",
+ " l2_regularization=0.005326508887463406,\n",
+ " learning_rate=0.060800813211425456, max_iter=512,\n",
+ " max_leaf_nodes=6, min_samples_leaf=5,\n",
+ " n_iter_no_change=5, random_state=1,\n",
+ " validation_fraction=None, warm_start=True)},\n",
+ " 13: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3afa53d60>,\n",
+ " 'cost': 0.028368794326241176,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3af75db20>,\n",
+ " 'ensemble_weight': 0.04,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3afa53a60>,\n",
+ " 'model_id': 13,\n",
+ " 'rank': 8,\n",
+ " 'sklearn_classifier': HistGradientBoostingClassifier(early_stopping=False,\n",
+ " l2_regularization=1.0647401999412075e-10,\n",
+ " learning_rate=0.08291320147381159, max_iter=512,\n",
+ " max_leaf_nodes=39, n_iter_no_change=0,\n",
+ " random_state=1, validation_fraction=None,\n",
+ " warm_start=True)},\n",
+ " 14: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3affae4f0>,\n",
+ " 'cost': 0.028368794326241176,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3af7a2340>,\n",
+ " 'ensemble_weight': 0.04,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3afc8fc10>,\n",
+ " 'model_id': 14,\n",
+ " 'rank': 9,\n",
+ " 'sklearn_classifier': MLPClassifier(activation='tanh', alpha=2.5550223982458062e-06, beta_1=0.999,\n",
+ " beta_2=0.9, hidden_layer_sizes=(54, 54, 54),\n",
+ " learning_rate_init=0.00027271287919467994, max_iter=256,\n",
+ " n_iter_no_change=32, random_state=1, validation_fraction=0.0,\n",
+ " verbose=0, warm_start=True)},\n",
+ " 16: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3ac655700>,\n",
+ " 'cost': 0.021276595744680882,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3ac1c60a0>,\n",
+ " 'ensemble_weight': 0.04,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3ac6552b0>,\n",
+ " 'model_id': 16,\n",
+ " 'rank': 2,\n",
+ " 'sklearn_classifier': HistGradientBoostingClassifier(early_stopping=True,\n",
+ " l2_regularization=3.387912939529945e-10,\n",
+ " learning_rate=0.30755227194768237, max_iter=128,\n",
+ " max_leaf_nodes=60, min_samples_leaf=39,\n",
+ " n_iter_no_change=18, random_state=1,\n",
+ " validation_fraction=None, warm_start=True)},\n",
+ " 17: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b09a1e50>,\n",
+ " 'cost': 0.03546099290780147,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3b02b41f0>,\n",
+ " 'ensemble_weight': 0.02,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b072a580>,\n",
+ " 'model_id': 17,\n",
+ " 'rank': 15,\n",
+ " 'sklearn_classifier': HistGradientBoostingClassifier(early_stopping=True,\n",
+ " l2_regularization=0.4635442279519353,\n",
+ " learning_rate=0.09809681787962342, max_iter=512,\n",
+ " max_leaf_nodes=328, min_samples_leaf=2,\n",
+ " n_iter_no_change=2, random_state=1,\n",
+ " validation_fraction=None, warm_start=True)},\n",
+ " 18: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b0782700>,\n",
+ " 'cost': 0.03546099290780147,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3b045f0d0>,\n",
+ " 'ensemble_weight': 0.02,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b09a1610>,\n",
+ " 'model_id': 18,\n",
+ " 'rank': 16,\n",
+ " 'sklearn_classifier': RandomForestClassifier(criterion='entropy', max_features=3, n_estimators=512,\n",
+ " n_jobs=1, random_state=1, warm_start=True)},\n",
+ " 19: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3aff3ceb0>,\n",
+ " 'cost': 0.028368794326241176,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3af9b7310>,\n",
+ " 'ensemble_weight': 0.02,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3aff3ccd0>,\n",
+ " 'model_id': 19,\n",
+ " 'rank': 10,\n",
+ " 'sklearn_classifier': ExtraTreesClassifier(criterion='entropy', max_features=448, min_samples_leaf=2,\n",
+ " min_samples_split=20, n_estimators=512, n_jobs=1,\n",
+ " random_state=1, warm_start=True)},\n",
+ " 20: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b13f13a0>,\n",
+ " 'cost': 0.07801418439716312,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3b0f96070>,\n",
+ " 'ensemble_weight': 0.02,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b13f10a0>,\n",
+ " 'model_id': 20,\n",
+ " 'rank': 23,\n",
+ " 'sklearn_classifier': PassiveAggressiveClassifier(C=0.14268277711454813, max_iter=32, random_state=1,\n",
+ " tol=0.0002600768160857831, warm_start=True)},\n",
+ " 22: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b018d370>,\n",
+ " 'cost': 0.028368794326241176,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3afc8f250>,\n",
+ " 'ensemble_weight': 0.04,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3affd3370>,\n",
+ " 'model_id': 22,\n",
+ " 'rank': 11,\n",
+ " 'sklearn_classifier': HistGradientBoostingClassifier(early_stopping=True,\n",
+ " l2_regularization=8.057778875694463e-05,\n",
+ " learning_rate=0.09179220974965213, max_iter=256,\n",
+ " max_leaf_nodes=200, n_iter_no_change=18,\n",
+ " random_state=1,\n",
+ " validation_fraction=0.14295295806077554,\n",
+ " warm_start=True)},\n",
+ " 23: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b0f28160>,\n",
+ " 'cost': 0.049645390070921946,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3b076e8b0>,\n",
+ " 'ensemble_weight': 0.06,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b0f15b20>,\n",
+ " 'model_id': 23,\n",
+ " 'rank': 20,\n",
+ " 'sklearn_classifier': MLPClassifier(alpha=0.02847755502162456, beta_1=0.999, beta_2=0.9,\n",
+ " hidden_layer_sizes=(123, 123),\n",
+ " learning_rate_init=0.000421568792103947, max_iter=256,\n",
+ " n_iter_no_change=32, random_state=1, validation_fraction=0.0,\n",
+ " verbose=0, warm_start=True)},\n",
+ " 24: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b12175e0>,\n",
+ " 'cost': 0.07092198581560283,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3b0eacd60>,\n",
+ " 'ensemble_weight': 0.06,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b1195910>,\n",
+ " 'model_id': 24,\n",
+ " 'rank': 22,\n",
+ " 'sklearn_classifier': RandomForestClassifier(criterion='entropy', max_features=16, n_estimators=512,\n",
+ " n_jobs=1, random_state=1, warm_start=True)},\n",
+ " 25: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b0d9c820>,\n",
+ " 'cost': 0.04255319148936165,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3b09421c0>,\n",
+ " 'ensemble_weight': 0.02,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3afab93d0>,\n",
+ " 'model_id': 25,\n",
+ " 'rank': 18,\n",
+ " 'sklearn_classifier': AdaBoostClassifier(base_estimator=DecisionTreeClassifier(max_depth=3),\n",
+ " learning_rate=0.046269426995092074, n_estimators=406,\n",
+ " random_state=1)},\n",
+ " 26: { 'balancing': Balancing(random_state=1),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b01788b0>,\n",
+ " 'cost': 0.028368794326241176,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3afeae8e0>,\n",
+ " 'ensemble_weight': 0.04,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b01784f0>,\n",
+ " 'model_id': 26,\n",
+ " 'rank': 12,\n",
+ " 'sklearn_classifier': ExtraTreesClassifier(criterion='entropy', max_features=414, min_samples_leaf=2,\n",
+ " min_samples_split=19, n_estimators=512, n_jobs=1,\n",
+ " random_state=1, warm_start=True)},\n",
+ " 27: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b0d6b9d0>,\n",
+ " 'cost': 0.04255319148936165,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3b0b09b80>,\n",
+ " 'ensemble_weight': 0.04,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b0d626d0>,\n",
+ " 'model_id': 27,\n",
+ " 'rank': 19,\n",
+ " 'sklearn_classifier': ExtraTreesClassifier(bootstrap=True, criterion='entropy', max_features=24,\n",
+ " min_samples_leaf=20, min_samples_split=5, n_estimators=512,\n",
+ " n_jobs=1, random_state=1, warm_start=True)},\n",
+ " 29: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b17611c0>,\n",
+ " 'cost': 0.14184397163120566,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3b12ac2e0>,\n",
+ " 'ensemble_weight': 0.06,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b154bdc0>,\n",
+ " 'model_id': 29,\n",
+ " 'rank': 25,\n",
+ " 'sklearn_classifier': GaussianNB()},\n",
+ " 30: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b15c7ee0>,\n",
+ " 'cost': 0.07801418439716312,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3b10e40a0>,\n",
+ " 'ensemble_weight': 0.02,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b159f4f0>,\n",
+ " 'model_id': 30,\n",
+ " 'rank': 24,\n",
+ " 'sklearn_classifier': AdaBoostClassifier(algorithm='SAMME',\n",
+ " base_estimator=DecisionTreeClassifier(max_depth=7),\n",
+ " learning_rate=0.02920859465461962, n_estimators=96,\n",
+ " random_state=1)},\n",
+ " 32: { 'balancing': Balancing(random_state=1, strategy='weighting'),\n",
+ " 'classifier': <autosklearn.pipeline.components.classification.ClassifierChoice object at 0x2ad3b107ad30>,\n",
+ " 'cost': 0.06382978723404253,\n",
+ " 'data_preprocessor': <autosklearn.pipeline.components.data_preprocessing.DataPreprocessorChoice object at 0x2ad3adf57af0>,\n",
+ " 'ensemble_weight': 0.04,\n",
+ " 'feature_preprocessor': <autosklearn.pipeline.components.feature_preprocessing.FeaturePreprocessorChoice object at 0x2ad3b107a970>,\n",
+ " 'model_id': 32,\n",
+ " 'rank': 21,\n",
+ " 'sklearn_classifier': ExtraTreesClassifier(criterion='entropy', max_features=1, min_samples_leaf=17,\n",
+ " min_samples_split=4, n_estimators=512, n_jobs=1,\n",
+ " random_state=1, warm_start=True)}}\n"
+ ]
+ }
+ ],
"source": [
+ "# Show the different models\n",
"pprint(automl.show_models(), indent=4)"
]
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 6,
"id": "027039cd",
"metadata": {},
- "outputs": [],
+ "outputs": [
+ {
+ "name": "stdout",
+ "output_type": "stream",
+ "text": [
+ "Accuracy score: 0.951048951048951\n"
+ ]
+ }
+ ],
"source": [
+ "# Predict the test labels\n",
"predictions = automl.predict(X_test)\n",
"print(\"Accuracy score:\", sklearn.metrics.accuracy_score(y_test, predictions))"
]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 13,
+ "id": "acd372ea",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "# Export the model with the highest rank\n",
+ "clf = automl.show_models()[7]['sklearn_classifier']\n",
+ "pickle.dump(clf,open('model.pickle','wb'))"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "a3324782",
+ "metadata": {},
+ "outputs": [],
+ "source": []
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "021b7159",
+ "metadata": {},
+ "outputs": [],
+ "source": []
}
],
"metadata": {