Turns out, my old Python assignments had plenty left to teach me.
Machine learning is often used to predict future values from data. But what if, instead of just predictions, you could express the relationships behind the data as a 'mathematical formula'?For example ...
The first bug its tests caught was the AI's own.