Due to recent updates, all users are required to create an Altair One account to login to the RapidMiner community. Click the Register button to create your account using the same email that you have previously used to login to the RapidMiner community. This will ensure that any previously created content will be synced to your Altair One account. Once you login, you will be asked to provide a username that identifies you to other Community users. Email us at Community with questions.

Performance output - what is the +/- value?

Rachel_dvmRachel_dvm Member Posts: 2 Learner I
in Help
Hello,

I have a quick question to which I couldn't find the answer by searching the other threads or the information available on the RapidMiner website. I'm not from a data science background but have been using RapidMiner for some classification using a few different machine learning processes. I have used cross validation and an example of the performance results is this (a very poor classifier for my data):

Precision: 95.06% +/- 6.06% (micro average: 94.76%) (positive class: true)

                                    True false         True true         Class precision
Pred. false                   3915                 587                  86.96%
Pred. true                    12                      217                  94.76%
class recall                  99.69%              26.99%       

I'm not concerned with the actual performance of the model but more the value I've bolded. What does this +/- range represent? I had originally thought it was a range of precision but the positive value is above 100%. Is this due to the class distribution difference?

Thank you in advance!

Best Answer

  • yyhuangyyhuang Administrator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 364 RM Data Scientist
    Solution Accepted
    Hi @Rachel_dvm,

    Suppose you have a 10 fold cross validation for your ML model, you will get 10 individual testing performances on each testing set. The  number after +/- is for the standard deviation out of the 10 testing performance. With 10 fold cross validated models, you will get the average performance as well. The average is the number before +/-.

    Cheers,
    YY

Answers

  • Rachel_dvmRachel_dvm Member Posts: 2 Learner I
    Yes, that is right. I used 10 fold. Thank you, that makes sense! Is there anything you could refer me to read to confirm this? (I will need to convince my supervisor that it is the standard deviation)
  • MartinLiebigMartinLiebig Administrator, Moderator, Employee, RapidMiner Certified Analyst, RapidMiner Certified Expert, University Professor Posts: 3,531 RM Data Scientist

    Best,
    martin

    - Sr. Director Data Solutions, Altair RapidMiner -
    Dortmund, Germany
Sign In or Register to comment.