-
Hi guys, Thanks for the fantastics package! It is really intuitive to use and the example notebooks are very helpful to get started. I have a question -- is it possible to combine multiple constraints (e.g., DemographicParity and EqualizedOdds) together when using say Exponentiated Gradient method? I did some search but was not able to find any documentation on this; is it advisable to subclass Thank you! |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
Thanks for kind words! We currently do not support what you're describing, but technically this is definitely possible. You could try to create your own moment by combining the code from I think technically cleanest would be to subclass
|
Beta Was this translation helpful? Give feedback.
Thanks for kind words! We currently do not support what you're describing, but technically this is definitely possible.
You could try to create your own moment by combining the code from
EqualizedOdds
andDemographicParity
, but that might not be the easiest route (you'd probably need to extend the functionality ofUtilityRateParity
to work with events that are not mutually exclusive).I think technically cleanest would be to subclass
fairlearn.reductions.ClassificationMoment
into a new class called something likeIntersect
(or perhaps a better name), which would be constructed by invokingIntersect(moment1, moment2)
, wheremoment1
andmoment2
would be arbitrarily initialized instances off…