I finished “taking” this short course yesterday. It felt to me like a decent introduction to supervised learning. It has whet my appetite and now I want to find another course.
I say “taking” because I skipped all the programming exercises, and towards the end when it started getting into the nitty gritty I started skipping whole sections.
My goal isn’t to become a machine learning engineer. I just want to understand machine learning at a high level so I can think about high-level things like self-driving cars (and the companies competing to deploy them) and artificial general intelligence.
This is analogous to how the typical machine learning engineer doesn’t need to understand much of anything about chip design. As long as the GPUs/CPUs/ASICs work, you don’t need to understand how they work. You don’t need to learn how to design a chip before you can be an effective machine learning engineer. You can abstract away the lower level details.
When I learn about machine learning stuff, I look for anything that looks like the equivalent of what a GPU/CPU/ASIC is for a machine learning engineer. Some low-level thing I can abstract away and treat as a black box. You can’t do that if you want to build these systems. But you can if you just want to think about high-level things.