Adding Global Actions in A Nutshell For Counterfactual Explainability (GLANCE) framework #196
+5,930
−0
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Outdated suggestions cannot be applied.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.
This PR is to add GLANCE, Global Actions in A Nutshell For Counterfactual Explainability framework.
GLANCE is a versatile and adaptive framework for generating global counterfactual explanations. These explanations are expressed as actions that offer recourse to large population subgroups. The framework aims to provide explanations and insights, ensuring that the actions benefit as many individuals as possible.
GLANCE comprises of two algorithms:
C-GLANCE: employs a clustering-based approach for generating global counterfactual explanations that allows the user to specify the number of global actions generated and serves as a tool to explain and debug ML models.
T-GLANCE: generates global counterfactual explanations using a decision tree-like structure. T-GLANCE's key strength is that allows policy-makers or auditors to select the features that determine the partitioning of the tree. By allowing
users to guide the feature selection, T-GLANCE can adapt to specific policy needs or audit requirements.
algorithm folders are here
example notebook can be found here
algorithm's dependencies are here
Closes #195 issue.