0
$\begingroup$

I see the authors of this paper are measuring F1 and NMI scores to measure the clustering quality. However, I don't understand the algorithm of how they actually measure it. See the Evaluation Section.

Although I have looked at the code, I am not sure about the actual algorithm.

$\endgroup$
0
$\begingroup$

Use the original papers, and books, on F1 and NMI.

There is little reason to only use a 2016 arxiv (non-reviewed) paper that uses them, because they have been introduced long before.

I don't see F1 mentioned anywhere in the paper,and it would be an uncommon choice for clustering (see literature for details).

$\endgroup$
  • $\begingroup$ N-Pair Loss facility location problem Lifted structure loss are also using similar kind of evaluation measures. That is the only reason I wanted to know. $\endgroup$ – Jay Patel Dec 31 '17 at 22:50
  • $\begingroup$ Yes, these are standard. Use a textbook, and/or the original sources, not some 2016 preprint that just uses them. $\endgroup$ – Anony-Mousse Dec 31 '17 at 23:38

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.