returning models used in adaboost python

Question

After applying adaboost on svm I want to know the models(their parametes) used in the adaboost algorithm.

ada=AdaBoostClassifier(n_estimators=10, base_estimator=SVC(probability=True)) 
ada.fit(x_train,y_train)

How can I find the models used in the adaboost.Thank you

Answer 1

The estimators_ field of your AdaBoostClassifier object contains each of your models. Viewing the details of those models will depend on what was used to build them. So for example you may need to view how to lookup how to get information for a DecisionTreeClassifier in the example below:

>>> from sklearn.datasets import load_iris
>>> from sklearn.ensemble import AdaBoostClassifier
>>> 
>>> iris = load_iris()
>>> clf = AdaBoostClassifier(n_estimators=2)
>>> clf.fit(iris.data, iris.target)
AdaBoostClassifier(algorithm='SAMME.R',
          base_estimator=DecisionTreeClassifier(compute_importances=None, criterion='gini',
            max_depth=1, max_features=None, min_density=None,
            min_samples_leaf=1, min_samples_split=2, random_state=None,
            splitter='best'),
          learning_rate=1.0, n_estimators=2, random_state=None)
>>> clf.estimators_
[DecisionTreeClassifier(compute_importances=None, criterion='gini',
            max_depth=1, max_features=None, min_density=None,
            min_samples_leaf=1, min_samples_split=2, random_state=None,
            splitter='best'), DecisionTreeClassifier(compute_importances=None, criterion='gini',
            max_depth=1, max_features=None, min_density=None,
            min_samples_leaf=1, min_samples_split=2, random_state=None,
            splitter='best')]
>>> 
>>> #first model
... clf.estimators_[0]
DecisionTreeClassifier(compute_importances=None, criterion='gini',
            max_depth=1, max_features=None, min_density=None,
            min_samples_leaf=1, min_samples_split=2, random_state=None,
            splitter='best')
>>> #second model
... clf.estimators_[1]
DecisionTreeClassifier(compute_importances=None, criterion='gini',
            max_depth=1, max_features=None, min_density=None,
            min_samples_leaf=1, min_samples_split=2, random_state=None,
            splitter='best')