Isn't every semi-supervised problem also a partially supervised learning problem and vice versa?


closed as unclear what you're asking by Sycorax, mdewey, mkt, kjetil b halvorsen, Michael Chernick Sep 15 '18 at 17:08

Please clarify your specific problem or add additional details to highlight exactly what you need. As it's currently written, it’s hard to tell exactly what you're asking. See the How to Ask page for help clarifying this question. If this question can be reworded to fit the rules in the help center, please edit the question.

  • 1
    $\begingroup$ I'm not deeply familiar with either term, but this issue seems to turn on definitions since the plain meanings of "semi-" and "partially" are synonymous. Can you find and provide definitions of both terms? $\endgroup$ – Sycorax Sep 14 '18 at 19:50
  • $\begingroup$ Semi-supervised learning is standard terminology, but I'm not familiar with the term "partially supervised learning". It has only 780 hits on google scholar (compared to 62,000 for "semi-supervised learning"). Just briefly perusing, many of them seem to go on to talk about semi-supervised learning (even using that terminology). $\endgroup$ – user20160 Sep 14 '18 at 22:46
  • $\begingroup$ Wiki says that pu-learning is semi-supervised learning where only positive instances are labeled. $\endgroup$ – mlvalidated Sep 15 '18 at 15:21
  • $\begingroup$ Yes, it's true that "PU" stands for "positive/unlabeled" but what does that have to do with your question? $\endgroup$ – Sycorax Sep 15 '18 at 19:27
  • $\begingroup$ I do not find very much other than that. I will just use the term semi-supervised learning. PU-learning doesn't seem popular. $\endgroup$ – mlvalidated Sep 15 '18 at 21:59