Summary
After reading this chapter, you should know how to leverage anchors, to understand the decision rules that impact a classification, and counterfactuals, to grasp what needs to change for the predicted class to change. You also learned how to assess fairness using confusion matrices and Google's WIT. Lastly, we covered CEM to explain a decision by what is minimally present and absent. In the next chapter, we will study interpretation methods for Convolutional Neural Networks (CNNs).