Options

Free community version

kksishtlakksishtla Member Posts: 2 Contributor I
edited November 2018 in Help

Hi all,

 

I am new to machine learning. I've recently downloaded the free version of RapidMiner. I am trying to figure out the different features. A couple of days ago I started running backward elimination but somehow it doesn't seem to work I mean it takes forever to run (I used the same operators as in the tutorial process flow).

 

Is it because the free version has a limitation of 10,000 data rows? My data file has 100K rows and 200 columns and I've taken a 50% random sample from it. 

 

Do I have to get a licensed version for this to work?

 

Thanks,

K

Answers

  • Options
    MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,517 RM Data Scientist

    Hi kkshishtla,

     

    which operator do you use inside the Backwards Elem? Some Machine learning operators run in O(N**3) so 10x the data means 1000x the runtime. 100.000x200 is to be honest quite a lot. I had a project with 80.000x180 and my neural net needed 6h. The new deep learning would be faster, but needs a license to get the full speed.

     

    ~Martin

    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • Options
    kksishtlakksishtla Member Posts: 2 Contributor I

    Thanks, Martin.

     

    I used the k-NN operator. 

     

    Does the issue have to do anything with the free version limitation of #10,000 data rows?

     

    Thanks,

    K

  • Options
    IngoRMIngoRM Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, Community Manager, RMResearcher, Member, University Professor Posts: 1,751 RM Founder

    Hi,

     

    Not with the row restriction, but the commercial version also offers parallel computing for some powerful learners which might speed up things considerably (depending on your hardware of course).

     

    Best,

    Ingo

Sign In or Register to comment.