Classification evaluation metrics: what do they really ask?

I’m quite sure that you are familiar with the foundational classification evaluation metrics and how to correctly calculate them for a given machine learning classification system (you would use an external python library, i know…). But, how confident are you that you really understand their meaning?

If you were to take a look into the pages of a typical technical textbook (or web page) for a description of the most common classification evaluation metrics you would find most approaches would choose to describe mathematical…