search

mmschlk / shapiq

Shapley Interactions for Machine Learning
https://shapiq.readthedocs.io
MIT License
60 stars 6 forks source link

add a baseline conditional imputer #154

Closed hbaniecki closed 1 month ago

hbaniecki commented 1 month ago

there are a few todos to be discussed f2f before the implementation

notes:

  1. ConditionalImputer requires x (as opposed to MarginalImputer), because value_function uses a distribution conditioned on x
  2. MarginalImputer.value_function() can be potentially improved to remove the for loop (like in ConditionalImputer)

progress towards https://github.com/mmschlk/shapiq/issues/40

mmschlk commented 1 month ago

Very cool. Will merge this for now and will update the repo code quality and tests on the weekend.