The Altair Community is migrating to a new platform to provide a better experience for you. The RapidMiner Community will merge with the Altair Community at the same time. In preparation for the migration, both communities are on read-only mode from July 15th - July 24th, 2024. Technical support via cases will continue to work as is. For any urgent requests from Students/Faculty members, please submit the form linked here.
Options

"How to use ensemble feature selection"

dina_alliasdina_allias Member Posts: 2 Contributor I
edited June 2019 in Help
Dear all,
Good day. I am a new user of rapid miner. By the way, I am looking for a way on how to use ensemble feature selection for text processing.
My plan is to combine three types of feature selection algorithm. But I'm not sure whether I can put all the three algorithms together into the ensemble feature selection or not.

Based on my understanding, if I want to use an ensemble method for the classifier, I can use voting, stacking or bagging method. But how about for feature selection ???. Really appreciated if anyone of you could show or teach me on how to use the ensemble feature selection model.

Thank you very much

Regards

Answers

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

    before going into the technical details on how to use RapidMiner to do the task, let me ask how exactly you want to combine three feature selections from an algorithmic point of view?

    Best regards,
    Marius
  • Options
    dina_alliasdina_allias Member Posts: 2 Contributor I
    Hi,
    Sorry for the late reply. My plan is to weight every feature from the feature selection algorithms. Then, from the result of the feature weighting, the best features will be used as an input for the classifier. But still wondering whether I can connect directly the feature selections together or not

    Thank you

    regards
  • Options
    MariusHelfMariusHelf RapidMiner Certified Expert, Member Posts: 1,869 Unicorn
    And what happens if the feature selection algorithms deliver contradictory weights? And how do you want to "connect" them (from an algorithmic point of view, not how to do it in RapidMiner).

    Best regards,
    Marius
Sign In or Register to comment.