Balanced package classifier

The accuracy of the classifier is the total number of correct predictions from the classifier divided by the total number of predictions. This may be good enough for a well-balanced class, but it's not ideal for the problem of an unbalanced class. Other indicators, such as accuracy, are a measure of how accurate the classifier's prediction is for a particular class, and recall is a measure of a classifier's ability to identify a class.

For an unbalanced F1 class data set, the result is a more appropriate indicator. This is the average harmonic precision and recall and the expression is -

from imblearn.ensemble import BalancedBaggingClassifier
from sklearn.tree import DecisionTreeClassifier
#Create an instance
classifier = BalancedBaggingClassifier(base_estimator=DecisionTreeClassifier(),
                                sampling_strategy='not majority',
                                replacement=False,
                                random_state=42)
classifier.fit(X_train, y_train)
preds = classifier.predict(X_test)

If you are looking for  an online  machine learning course  then check this out