Explainability: The Last Mile – Towards Data Science

For your user to understand your model it's not enough for it to be'explainable' -- you need to provide the ultimate explanation Interpretable or explainable models have gone from being almost a chimera to being an increasingly common business as usual requirement. However, although there are a growing number of methods available to explain models, these are still technical tools, that are aimed at statistical and data science practicioners who need to understand the models they create. They are a necessary, but insufficient step towards creating models that are understandable by the end user. Creating a model that an end user can understand means on the one hand ensuring that they understand at a basic level what the input and output variables in the model are, and on the other that they understand how those variables operate within the model. In each of these cases the final presentation is crucial to ensuring the end goal of a seamless user experience is met.