Hitachi Vantara Pentaho Community Forums
Results 1 to 2 of 2

Thread: visualize classifier errors

  1. #1
    Join Date
    Dec 2015
    Posts
    1

    Default visualize classifier errors

    Good evening everyone,

    It's maybe a dumb question, anyway... How can I know which item is badly predicted by my Weka classifier? How can I do that without involving filename string attribute in my training set (which is a useless feature in order to build the model) ??

    Thanks in advance,
    Laszlo

  2. #2
    Join Date
    Aug 2006
    Posts
    1,741

    Default

    You can add an ID field into your data (either using one that's in your data already, or by using the AddID filter). Then use the FilteredClassifier along with your base classifier and the Remove filter (to remove the ID at training and prediction time). This way the base classifier doesn't use the ID for learning the model, but it stays in the data for visualization/output of predicted instances etc.

    Cheers,
    Mark.

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.