The kind of course I’d lead in recent times, for adults (or I suppose precocious youngsters bored with the conventional curriculum for their age), was what you might call “AI for janitors”. By that I mean we’d learn the data harvesting and clean up steps prior to siccing one of the divinatators on it. By “divinator” I mean “one that divines” by means of specific algorithms we’d pull up inside of sci-kit learn (sklearn).
Divinators: random forest, K nearest neighbors, support vectors (multidimensional, hyper), and of course whatever deep or shallow neural nets. But these data devouring beasts require “clean” food, meaning you’re better off not feeding them raw data. There’s a lot of “food prep” that needs to go on, and that’s where I’d come in with my Jupyter Notebooks and so on.
However, going back to Pythonic basics (Pythonica), I’m not averse to barking up a different branch, meaning SQL roots might take you towards website development (back end mostly still, if using Python), which involves entirely different skill sets in many cases, or at least until recently.
I’m showing this “river delta” or “tree” pattern above, which is how entropy fans out and burns calories, to paraphrase Adrian Bejan. Up the left side (house left) we see the web developer namespaces, such as Django (named for the gypsy) and Flask. Up the right side: data science, which was numpy + pandas in my day and age.
What’s happening more recently is data visualization world is colliding with web-displayed dashboard world (remember the Johns Hopkins dash during covid?), so you end up learning HTML / CSS either way, and the JavaScript DOM. JS tends to be less about data science and more about web development, whereas Python frequents both sides of that membrane.
I’m not saying you can’t do data science in JavaScript, and with my Jupyter Notebook solution (the solution I favor, I’m not a developer on that project) we have JS built right in, to use in combo with our Python kernel.
Am I leading such an “AI for janitors” course today? Nope. I’m barking up the web developer side focusing on Flask, in a demo for my Supermarket Mathematics. In the meantime, we study Fred Meyer (the local Krogers) from an intellectual historical point of view. Meyer Memorial Trust is a former client of mine actually, although I haven’t gotten into the FreeGeek chapter yet, in my Winter Term chronology.
