Python Ensemble VotingClassifier for regression

Question

I want to build an Ensemble method to predict sales. Suppose I use estimators Linear Regression , Decision Tree Regression and ARIMA .

I decided to take the mean of the above estimators to be my predicted value. Can I use the VotingClassifier class to build this ensemble? or does it work only for classification problems?

Is there some other library in python for combining different regression estimators (not classification and it can't be the same estimator like in bagging regression)

Answer 1

Simply make predictions and take an average of them.

https://scikit-learn.org/stable/modules/ensemble.html#voting-classifier

For classification there are two ways:

hard voting - class which was predicted by most models is selectes;

soft voting - each model predicts probabilities, classes with the highest probability are chosen.