Options

How to get confusion matrix when carrying out classification cross-validation

gwingwin Member Posts: 4 Newbie
edited October 2021 in Help
Hello,

I have been trying to get my confusion matrix from when carrying out K means cross validation for my data set but it has been abortive. Can anyone help?
I have attached my process below.
Thank you.

Best Answers

  • Options
    MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,508 RM Data Scientist
    Solution Accepted
    Hi,
    try to turn off the parallelization of cross validation. Then you can see the last iteration's value.

    Best,
    Martin
    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • Options
    gwingwin Member Posts: 4 Newbie
    Solution Accepted
    Hello,

    It worked!

    Thanks a lot for this.

    Best Regards,

    Godwin

Answers

  • Options
    gwingwin Member Posts: 4 Newbie
    hello,

    So sorry I mixed things up, what I was doing was classification matrix (I later did K-Means but not for this process). But my question is why I could my cross-validation tool (during classification) not carry out my test and training set(at the same time) and bring out a confusion matrix table. When I tried looking for the performance vector (in my testing panel) from the performance tool, it seem not to bring out the performance vector option so that I can see the confusion matrix. (I hope this is clear).

    I have attached the process again.

    Best Regards,

    Godwin
  • Options
    MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,508 RM Data Scientist
    Hi,
    maybe just because you did not connect the per port of x-validation to the right hand side?

    Cheers,
    Martin
    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
  • Options
    gwingwin Member Posts: 4 Newbie
    Hello,

    Thanks for the speedy response.
    And yes that works for me if I want to get only the confusion matrix but then my professor's version can get the matrix even when he connects mod (instead of per) to res (and then right-click the performance inside the cross-validation to see the performance vector but I can't see it in mine when I follow the same steps).

    I hope it is clear.

    Godwin
Sign In or Register to comment.