Data Science Stack Exchange is a question and answer site for Data science professionals, Machine Learning specialists, and those interested in learning more about the field. It only takes a minute to sign up.
Sign up to join this community
What is the difference between interpretable algorithm and transparent algorithm? In particular, is there an algorithm that is interpretable but not transparent?
Update: In this post we defined explainability and interpretability and showed how these two concepts are different. The question is whether transparency is something different from either (or both) explainability or interpretability.
In the this link, the author describes that the difference between model transparency and explainability (ie interpretability) lie in their approaches to solving the black box problem in AI.
Transparency solves this by the use of simple models, such as linear regression, that are well-known and use easy to follow steps. Whereas interpretability solves this problem by trying to gain insights into the black box using statistical methods.
The most popular statistical method is the Local Interpretable Model-Agnostic Explanations framework (LIME). This is where, quote,
This framework was suggested and further explained in this paper that also came with a python package called LIME.
Using the above framework to answer your second question, Deep Neural Networks can be made explainable (ie interpretable ) without necessarily being transparent.
