Created
August 21, 2018 12:21
-
-
Save humamfauzi/31cdd60ea9f9c94e4b11f6bbe8645477 to your computer and use it in GitHub Desktop.
Using Generated Data and Ensemble Learning in Python sklearn
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
# Please use Jupyter notebook to test this. | |
from sklearn.datasets import samples_generator | |
from sklearn.tree import DecisionTreeClassifier | |
from sklearn.ensemble import AdaBoostClassifier, BaggingClassifier, VotingClassifier | |
from sklearn.model_selection import StratifiedKFold, GridSearchCV, train_test_split | |
import numpy as np | |
import pandas as pd | |
X, y = samples_generator.make_classification(n_samples = 5000, | |
n_features = 20, | |
n_informative = 13, | |
n_redundant = 0, | |
n_classes = 2, | |
n_clusters_per_class = 2) | |
X, y = pd.DataFrame(X), pd.DataFrame(y) | |
rstate = 43 | |
kf = StratifiedKFold(n_splits=10, shuffle=True) | |
dtc = DecisionTreeClassifier() | |
dtc_params = {} | |
dtc_params['max_depth']=[19] | |
dtc_params['max_features']=[15,18] | |
dtc_params['min_samples_leaf']=[6] | |
dtc_params['min_samples_split']=np.logspace(-3,0,10) | |
dtc_params['random_state']=[rstate] | |
dtcGS = GridSearchCV(dtc, dtc_params, cv=kf, scoring='accuracy', verbose=1) | |
dtcGS.fit(X, y) | |
print(dtcGS.best_score_) | |
abc = AdaBoostClassifier(dtcGS.best_estimator_, n_estimators=200) | |
x_train, x_test, y_train, y_test = train_test_split(X, np.array(y).reshape(5000), test_size=0.2, random_state=rstate) | |
abc.fit(x_train, y_train).score(x_test, y_test) | |
# 0.95099999999999996 | |
bc = BaggingClassifier(dtcGS.best_estimator_, n_estimators=200, random_state=rstate) | |
bc.fit(x_train, y_train).score(x_test, y_test) | |
# 0.90600000000000003 | |
vc = VotingClassifier(estimators=[('adaBoost',abc), ('bagging', bc)], voting='soft') | |
vc.fit(x_train, y_train).score(x_test, y_test) | |
# 0.94799999999999995 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment