Hitachi Vantara Pentaho Community Forums
Results 1 to 2 of 2

Thread: Why am I getting a 1.000 ROC area value even when I don't have 100% of accuracy?

  1. #1
    Join Date
    Aug 2016

    Default Why am I getting a 1.000 ROC area value even when I don't have 100% of accuracy?

    Sorry if this issue has been posted before, but I couldn't find it

    I am using Weka as a classifier, and it has worked great for me so far. However, in my last test, I got a 1.000 ROC area value (which, if i remember correctly, represents a perfect classification) without having 100% of accuracy, as can be seen in the Confusion Matrix in the Figure.

    My question is: Am I interpreting the results incorrectly or am I getting wrong results (maybe the classifier I am using is badly programmed, although I don't think it's likely)?

    I've tested it in versions 3.6 and 3.8 and had the same problem in both of them.

    Thank you!

  2. #2
    Join Date
    Aug 2006


    ROC is a summary statistic that measures the ranking performance of a classifier. Accuracy is just one point on the ROC curve - the point that corresponds to a threshold of 0.5 on the probability assigned to the positive class (in a two class problem). If you imagine a classifier that always ranks the positive class higher than the negative one, when sorting predictions according to the probability assigned to the positive class, then this classifier will have an AUC of 1.0. If some negative instances receive a predicted probability of being positive that is greater than 0.5 (but still less than the probability assigned to true positive instances) then this will result in misclassification when using a threshold of 0.5.


Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
Privacy Policy | Legal Notices | Safe Harbor Privacy Policy

Copyright © 2005 - 2019 Hitachi Vantara Corporation. All Rights Reserved.