Create binary and multiclass objective classes #504
Conversation
* Update latest dependencies
* adding functionality and test * changelog and adding boolean param
* Add a whitelist for update_deps check * Remove from expected * Update deps * Changelog
… into 373_objective_classes
|
|
||
| def decision_function(self, y_predicted, extra_cols, threshold): | ||
| """Determine if transaction is fraud given predicted probabilities, dataframe with transaction amount, and threshold | ||
| def decision_function(self, ypred_proba, threshold=0.0, X=None): |
There was a problem hiding this comment.
This is great.
Thinking out loud, I wonder if in the future we'll want to have the threshold default to None here and throw an error, to indicate to users that they must always decide on a threshold.
There was a problem hiding this comment.
LGTM!! Nice work shepherding this through! This is huge :D and very close to getting into master!
The one final thing I'd like before you merge is that let's delete component_graph = [] from ClassificationPipeline as that could cause bugs. Other than that, all the important comments can go be addressed in a separate PR or filed as an issue.
I haven't finished going through the tests, but since this is merging to a feature branch, let's merge now and address test comments in a separate PR too.
|
Ok cool I see your latest updates now. Good to merge from my perspective! 🥳 |
* Objectives API: Create new binary / multiclass pipeline classes and remove objectives from pipeline classes (#405) * Objectives API: Remove ROC and confusion matrix as objectives (#422) * Change `score` output to return one dictionary (#429) * Create binary and multiclass objective classes (#504) * Update dependencies (#412) * Hide features with zero importance in plot by default (#413) * Update dependencies check: package whitelist (#417) * Add fixes necessary for docs to build for improved objectives project (#605) * Remove calculating plot metrics from AutoML (#615)
Continuation of #419 which was closed after a force push 😅(Closes #504, #488)
As part of #346, we need to split our objectives into binary and multiclass objectives.