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.

"I doubt the accuracy is calculated right?"

siamak_wantsiamak_want Member Posts: 98 Contributor II
edited June 2019 in Help
Hi the international RM team,

I run a neural network binary classification process which the performance is evaluated with an X-Validation operator (10-fold). I receive an astonishing result:
the accuracy is : 100%, but the precision = 97.2% and also the recall=97.2%!!!

The equation for calculating accuracy is: (TP+TN)/ (TP+TN+FP+FN)
the precision can be calculated as : TP/TP+FP; So if the accuracy =100%. then FP=0 and consequently, precision=100.

Do I misunderstand something? Please explain me. any help to eliminate my doubt would be appreciated.

Thanks.
Tagged:

Answers

  • MariusHelfMariusHelf RapidMiner Certified Expert, Member Posts: 1,869 Unicorn
    From your description something seems to be wrong. Can you please provide your process and the data?
  • siamak_wantsiamak_want Member Posts: 98 Contributor II
    Hi Marius,

    I think I found the problem, RM delivers "Weighted Mean Precision", I think this is not the precision which I wrote the equation for, in my previous post.  RM also delivers "mikro average" in parenthesis just infront of the "Weighted Mean Precision". The micro average adheres to the equation which I wrote. Would you please explain me which precision should I use? I mean weighted mean or the mikro one?

    Thanks again

Sign In or Register to comment.