AUPRC

DocMusher
DocMusher New Altair Community Member
edited November 2024 in Community Q&A
Who is able to help us.
It eems that we will have to implement AUPRC for the paper (because all papers in PLOS I are using this measure in case of inballanced data).

I tried to do it with R extension, and succeded to evaluate models, but I cannot use it in optimization and validation of models (because execute R cannot provide object "averagable" that is used in X-validation).

So I wanted to implement AUPRC as new operator, but I cannot find source code of RM6.5. Is it still open? and if it is where can I find it.

Or if you have some other way to include AUPRC in process please tell me.

Sven and Milan

Answers

  • Marco_Boeck
    Marco_Boeck New Altair Community Member
    Hi,

    as of Studio 6.5, the core has been released once more under AGPL3. The sources can be found by either clicking the "Source Download" link on the actual download page of RapidMiner Studio or by simply clicking here: https://github.com/rapidminer/rapidminer-studio

    Regards,
    Marco
  • MartinLiebig
    MartinLiebig
    Altair Employee
    Hi sven,

    i responded by mail, but for all of you:
    Extract Performance solves the problem. You can generate any performance value in R/Python and them convert the example set into an RM Performance vector with this operator.


    Yours,
    Martin
  • Milan_Vukicevic
    Milan_Vukicevic New Altair Community Member
    Thank you all

    This is solved