Interpretability and explainability of predictive models

https://freakonometrics.hypotheses.org/64629