Is it possible to use different classifiers in sklearn.ensemble?

Question

I have been learning sklearn.ensemble, for example, clf =AdaBoostClassifier(svm.LinearSVC(),n_estimators=10,max_samples=0.1,)

My question is whether I can use different classifiers as the base_estimator, because base_estimator can only accept one Object, if I want to use LogisticRegression() to continue classifying instances that are misclassified by LinearSVC(), what should I do? Is it possible?

Answer 1

AdaBoostClassifier can only train homogeneous estimators. scikit-learn does not support training heterogeneous ensembles. You might want to have a look at pyensemble.