[MRG] Documentation: introduction to metric learning #145
Conversation
|
I still need to write a few lines about use-cases. Note that I also commented out the part listing the class methods: I think it would be better to move this to the supervised and weakly supervised sections, within a proper 'API' subsection |
|
This should be ready to merge - feedback is welcome. |
Yes I agree, the current 'Package overview' subsection is currently not organized. We could add these comment lines, plus also organize the sections a bit (like separate weakly supervised algos from supervised etc). I'll raise an issue for that |
|
Merged, thanks! Moving the commented-out section should be handled by gh-149. |
This PR improves the documentation by adding a proper section introducing metric learning (problem setting, mahalanobis distance, use cases and general references). A small description of the package is also added to the index page for clarity.