Hitachi Vantara Pentaho Community Forums
Results 1 to 1 of 1

Thread: Multiclass Classifier - how are probabilites generated

  1. #1
    Join Date
    Oct 2017
    Posts
    1

    Default Multiclass Classifier - how are probabilites generated

    Hi,

    I'm using the MultiClass classifier with ECOC (Random codes) and LMT as base classifier.

    As far as I understand from what I read, the ECOC decomposes the problem in many binary ones (so many as the number of digits of the codeword).

    So given that the codeword is for example 5-digits then 5 LMT classifiers will be constructed and each one will predict the 0 or 1 value for the specific bit.

    Then the resulted codeword is compared to the codeword of each class based on Hamming distance and the one with the smallest one is given as the output of the classifier.

    So far so good. My question is how are the probabilities for each class that the MultiClassClassifier produces generated?
    Last edited by efismefi; 10-23-2017 at 08:11 AM.

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 - 2017 Pentaho Corporation. All Rights Reserved.