image

🎉 🎉 RAPIDMINER 9.10 IS OUT!!! 🎉🎉

Download the latest version helping analytics teams accelerate time-to-value for streaming and IIOT use cases.

CLICK HERE TO DOWNLOAD

Construct A New Tree from Random Forest Result

dragonedisondragonedison Member Posts: 17  Maven
edited November 2018 in Help
Dear everyone,

My dataset has 130 features, and I use one feature a time to train a classifier with Random Forest. Some of these 130 generated classifiers give good results. However, when I use two features in Random Forest(RF) training, the performance is not that good. So I would like to use several of the 1-feature-trained RF classifier to build up a tree classifier by giving the results of the 1-feature-trained classifiers different weights. I would like to know what operators can I use in RapidMiner to implement this. Do I have to work in the code level?

Thanks!
Gary

Answers

  • dragonedisondragonedison Member Posts: 17  Maven
    Dear everyone,

    I would like to make my request more obvious. Since there will be a confidence score for the Random Forest classification result. Are there any ways to combine different Random Forest classifiers' confidence scores to generate a new tree classifier?

    Thanks.
    Gary
  • landland RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 2,531   Unicorn
    Hi Gary,
    yes you can do this using the voting operator.

    Greetings,
      Sebastian

Sign In or Register to comment.