The wall has a name: linear algebra

I hit a wall in my quantum research. The wall has a name.

Linear algebra was, in my entire math education, approximately one day of high school. The lesson went something like: here are matrices, they're easy, moving on. I had no reason to push back on that at the time.

The traditional college-prep path (Algebra I, Geometry, Algebra II, Pre-Calc) doesn't go near it. In college I took stats and calc. Linear algebra just never came up. Apparently this is common, and apparently it matters enormously, because quantum computing runs on it.

The good news is that ML does too. So this detour is buying two things at once.

There's a known split in how linear algebra gets taught: the engineering version, which is computational and applied, and the mathematician's version, which is proof-heavy and abstract. The consensus I've seen is that it rewards abstract thinkers and punishes people who are expecting something like calculus. I'll find out.

My current stack: The Manga Guide to Linear Algebra (legit good; No Starch Press knows what they're doing), Linear Algebra for Dummies, and Khan Academy. As an aside, I noticed that Khan’s exercise sequences don’t really sequence well with the content. It’s satisfactory enough for me, though.

When I have some footing under me, I have Linear Algebra for Data Science, Machine Learning and Signal Processing, Math for Deep Learning, and Linear Algebra Done Right waiting (the last one being the proof-heavy mathematician's version), so I'm saving it for when I've earned it.

The standard curriculum runs roughly: systems of linear equations, matrices, determinants, vector spaces, linear transformations, eigenvalues and eigenvectors. I'm currently in vectors and should be at linear transformations soon.

I'll report back when I'm far enough in to have an opinion worth sharing.