Chapter 10. Metrics of Machine Learning
Young man, in mathematics you don’t understand things. You just get used to them.
—John von Neumann
At its core, machine learning is about devising, and then building, a software model that can predict a result given some input data. The key difference with any other kind of classic software is that the software behind a machine learning model doesn’t walk its way through a set of predetermined logical routes. The software behind a machine learning model is opaque in the sense that it gives a result, but the steps taken to produce it are not visible and interpretable as the source code of a canonical routine.
You can see a machine learning model as a mathematical function that produces output as a prediction, ...