fixout 0.1.22

Algorithmic inspection for trustworthy ML models

Algorithmic inspection for trustworthy ML models

• Website
• Documentation
• Blog

Getting started

How to start analysing a simple model (let's say you have trained a binary classifier on the German Credit Data):

from artifact import FixOutArtifact
from helper import FixOutHelper

fixout = FixOutHelper("Credit Risk Assessment") 

# Indicate the sensitive features
sensitive_features = [(19,0,"foreignworker"), 
                      (18,1,"telephone"), 
                      (8,2,"statussex")] 

# Create a FixOut Artifact with your model and data
fxa = FixOutArtifact(model=model,
                     training_data=(X_train,y_train), 
                     testing_data=[(X_test,y_test)],
                     features_name=features_name,
                     sensitive_features=sensitive_features,
                     dictionary=dic)

Then run the inspection

fixout.run(fxa) 

Finally, you can access the generated dashboard at http://localhost:5000 ;)

You should be able to see an interface similar to the following