I am trying to implement a feature into my machine learning model in which an entry feature is ranked in terms of other entry features.

So normally I would just have the feature ranked (1st...2nd....3rd) however its important to know how far each ranking is in relation to the other entries.

Student A Scores 100%
Student B Scores 85%
Student C Scores 50%
Student D scores 10%

For example, I have 4 separate entries above. So in my model instead of entering 10% for the last row as a feature for my model I would like some sort of mathematical or machine learning approach to enter a value that would show how far that 10% is in actuality from all the other 3 entries as a whole.

  • 2
    $\begingroup$ It sounds like you want to normalise the feature based on its range. Since your feature already is a percentage value ranging from 0 to 100 it already does what you need. Can you explain why 10% which is a fifth of 50% and this way representing the distance is not sufficient for your purposes? $\endgroup$ – Gegenwind Mar 24 '18 at 16:41

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.