Hitachi Vantara Pentaho Community Forums
Results 1 to 2 of 2

Thread: Possible Arff or Xrff format question

  1. #1
    Join Date
    Mar 2016
    Posts
    1

    Default Possible Arff or Xrff format question

    Hé guys,

    I'm wondering if the following is possible: so far, with some fiddling, I can't figure it out.
    I have several image with feature extraction from google inception. This means that all my images have 1000 features with a weight attributed to them. (Chance of it being in there)
    The top 5 weighted features are selected to be the only important ones, because the rest is a really low chance.
    The question is, is it possible to select AND weigh a feature based on the list of a 1000 features, or do I have to manually set all values to 0 except the top 5 values?
    I also do this in Multiple Instance learning, so that I somethings get 10,000 values on one line, combining that with a training set of 2000 creates a tidiously slow algorithm.
    I hope I was clear enough.

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

    Default

    I'm not sure that I understand. Can't you just remove all the features that have zero weight from the data entirely?

    I think the only algorithm in Weka that can make use of attribute weights directly is naive Bayes.

    Cheers,
    Mark.

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.