Options

"Process decision tree in RapidMiner 5.3"

facundo8facundo8 Member Posts: 2 Contributor I
edited June 2019 in Help
Hello everyone. I have a question. The method used by the process "decision tree" in RapidMiner 5.3 to generate the tree is C4.5?

Thank You.
Tagged:

Answers

  • Options
    sgenzersgenzer Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM Moderator Posts: 2,959 Community Manager
    hi...I am not sure about "Decision Tree" but fairly certain that the "W-J48" operator is the same as C4.5 (or that was what was explained to me in the Weka course).  If you download the Weka expansion pack, you will see it there.

    Scott
  • Options
    MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,507 RM Data Scientist
    I would recommend using WEKA there as well. I am not sure about the concrete algorithm behind the DecTree. The J48 from Weka is really a C4.5 (see http://weka.sourceforge.net/doc.dev/weka/classifiers/trees/J48.html ).
    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • Options
    facundo8facundo8 Member Posts: 2 Contributor I
    Hi.
    Yes, i have the weka extension but my problem continues because the image of the decision tree generated by the operator W-J48 is very unreadable due to the extension of the tree. I attach an imagen.
    I was recommended to make an process with "W-J48" operator to generate the rules and include another operator (i don't know which one!) that take the rules generated in W-J48 and generate an tree more "presentable". I need it because is my thesis.
    Hope you can understand me.

    Thank you so much.



    image
  • Options
    sgenzersgenzer Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM Moderator Posts: 2,959 Community Manager
    hi yes the graphical output of the W-J48 decision tree is pretty terrible.  I never use it.  Rather I use the "textual output" which is much better.  You can see it under the "description" tab in results.  It will look something like:

    W-J48

    J48 pruned tree

    ------------------

    edad <=37
    |    dependencia <=23
    |    sdsdsd
    |    sdsds

    etc...

    You simply read the tree from top to bottom, left to right.

    Scott
Sign In or Register to comment.