Learning anything involves a lot of experimentation, but in art the feel you get is very close to what you're actually doing, while in math and engineering the

Learning anything involves a lot of experimentation, but in art the feel you get is very close to what you're actually doing, while in math and engineering the "feel" is something you learn to interpret with experience applying a theoretical point of view to numerous problems. Even low-grade data science where cobbling together some grid-searched scikit-learn stuff suffices is more like fixing a damaged combustion engine than riding a bike.