Increasing max_iter to 1000 for LogisticRegression used in PrejudiceRemover #254
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
The
PrejudiceRemover
inprocessing bias mitigation algorithm usesLogisticRegression
, and by default,max_iter
forLogisticRegression
from sklearn is 100. However, in practice, more iterations are occasionally needed to fitPrejudiceRemover
to large datasets, some of which are common benchmarks in the fairness literature.With this in mind, this PR increases the
max_iter
threshold to 1000 from 100 in order to helpPrejudiceRemover
properly converge on large datasets.