So far, the plugin creates a bunch of evidence - but how to use it for an audit? Provide a sample Jupyter Notebook that demonstrates what can be read from the evidence.
[ ] optional: re-building the model using the given parameters and training & testing it in Jupyter notebook. Changing parameters to see if that improves any bias.
Provide the example analysis
[x] upload the notebooks & used evidence in a repo
So far, the plugin creates a bunch of evidence - but how to use it for an audit? Provide a sample Jupyter Notebook that demonstrates what can be read from the evidence.
Provide the example analysis