Due to recent updates, all users are required to create an Altair One account to login to the RapidMiner community. Click the Register button to create your account using the same email that you have previously used to login to the RapidMiner community. This will ensure that any previously created content will be synced to your Altair One account. Once you login, you will be asked to provide a username that identifies you to other Community users. Email us at Community with questions.

Weighted Voting Ensemble

mrfarhankhanmrfarhankhan Member Posts: 14 Contributor II
edited November 2018 in Help
Hello All,

I am doing a research on heterogeneous classifier ensembles to improve prediction accuracy. I have been searching to find weighted voting in rapid miner but unsuccessful. I have implemented simple majority voting using the Vote operator, but now I want to implement Weighted Voting. How can I do that ? If it's not built-in, then please let me know if any third party tool is available ?

Grateful
Farhan !

Answers

  • MariusHelfMariusHelf RapidMiner Certified Expert, Member Posts: 1,869 Unicorn
    Hi Farhan,

    which criteria should be used for weighting?

    Best regards,
    Marius
  • mrfarhankhanmrfarhankhan Member Posts: 14 Contributor II
    Marius wrote:

    which criteria should be used for weighting?
    Thanks for your reply. Weightage criteria can be accuracy or f-measure of individual classifiers used in the ensemble.
  • MariusHelfMariusHelf RapidMiner Certified Expert, Member Posts: 1,869 Unicorn
    Hi,

    unfortunately there is no single operator that can do it in one step. You have to create your own process for it. It will probably be relatively complex and include X-Validations to calculate the performance, joining together the results of the base learners  and then manually aggregating and weighting the predictions.
    What could be interesting is the Stacking approach, where you first create some base classifiers and then train a final model on the predictions of the base classifiers. It is expected that the final model will assign higher weights to those base classifiers that deliver better predictions.

    Best regards,
    Marius
Sign In or Register to comment.