I want to ensemble three or four regression model: like GBDT, XGBDT, SVM. I know there are votingC = VotingClassifier() for classifier. I want to know are there some methods or function for ensemble regression model. Now, I just know set the simple weight for each model. how about other good method for ensemble regression model by python. Thanks!


There are many ways in which you could create an ensemble from your base models. Some resources to take a look at are the following:

Firstly, I would point your attention towards this question, that has a lot of answers and interesting ideas.

Regarding implementation:

I have used in the past the brew module that has extensive implementations of different stacking-blending techniques etc, while using different combination rules and diversity metrics. It is also compatible with sklearn models.

Another interesting project is xcessiv, which I haven't used myself but provides a helpful gui for managing your ensemble's details.

Finally, regarding theoretical details, I would suggest you take a look into this survey that focuses on ensembles in regression tasks.

| improve this answer | |

Not the answer you're looking for? Browse other questions tagged or ask your own question.