🎉 🎉 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

"Clustering different models"

OzoneOzone Member Posts: 17  Maven
edited May 2019 in Help
Hi,

I create classification models for about 1000 datasets. Now I want to find similarities or clusters in these models.

My models are decision trees. So I am looking for an operator that compares features and tresholds used in my trees to build some clusters out of my 1000 trees?!

Could anyone help me with telling if this is possible and what is the operator?

Greetz
Ozone
Tagged:

Answers

  • Fabian_WewersFabian_Wewers Member Posts: 3 Contributor I
    Hi Ozone,

    may it be possible for you to log the threshold and features with the operator "Log"? Afterwards you could use the operator "Log to Data", create an exampleset and use every clustering method which rapidminer offers.
    I hope my comment will support your project :-)

    Greetz
    Fabian
  • OzoneOzone Member Posts: 17  Maven
    Thanks for your answer!

    I totally understand your advice but I don't think that there is any possibility to log a model tree structure (features & tresholds)....





  • awchisholmawchisholm RapidMiner Certified Expert, Member Posts: 458   Unicorn
    Hello

    You could write the model out as a PMML or XML file then read it back in and use XPATH to query the nodes.

    Not for the faint of heart.

    regards

    Andrew
Sign In or Register to comment.