Ask HN: I suck at math, where to start?

What do you want to learn? And what area do you want to work in?

From a CS view math is:

Discrete math and combinatorics? Really useful for proving algorithm properties etc.. graph theory is useful as well.

Statistics? Extremely useful. Linear algebra? Extremely useful as applied to statistics and deep learning.

Theorem proving? Useful for determining program correctness, important in some industries.

Of these, I think stats and linear algebra are the most fundamental. You can use these to build models of things and estimate parameters and create predictions.

I think the critical piece is to learn how to apply these tools/concepts correctly to solve problems, determine when they are valid / what the limits are / and how to intellectually debug them.

Otherwise learning about algorithmic complexity and how to solve CS type problems with algorithms is more likely to help your career.

If you have a math-savvy person in your circles, it might be a good idea to set up a chat with them. I think a good beginning point would be to identify a specific problem or concept that you've encountered that gave you some difficulty, and then talk it through with them.

In my experience as a mathematician, it is not always easy to identify the source of frustration in a problem. Having a different perspective can be really helpful. (Also in my experience as a math educator, a shockingly large proportion of problems stem from fractions, exponentials, and logarithms --- or not knowing what a function is.)

There's nothing better for filling math gaps than going through Kahnacademy.

Start skimming at the 1st grade level and then slow down and focus when you reach material that isn't easy.

Same. I definitely don't need math in my day-to-day. But I've been working on this side project for years, and haven't made much progress because I've hit a ceiling that requires math to transcend.

Thank you, but I am mathematically illiterate. I can't identify greek letters after sigma without looking them up, I ignore equations in papers and other texts unless they are supporting something that offends me, I don't have a working understanding of why certain operations are meaningful or their uses other than knowing there is probably a python library that does anything I could need.

However, things have shapes and relationships, most of which we can't physically see, but we can reason about and compare them, and that's about as deep as I get. e.g. do things at a certain level of abstraction have a similar shape, and what are the words that describe that? Does this tell us more about the thing itself, or the limits of our ability to percieve it, and it's just an artifact of our lens? It's like having an ear for music, where you hear fragments of other pieces in everything, and interpret forms and symmetries and respond to intuitions, expectations, and resolutions - but not being able to play it.

I think we're entering into an era where we can finally have punk math, where some idiots get on stage and the people watching them go, "omg, this is terrible but fun, and I could do this better," and a thousand bands get launched. When I write stuff like that, it's because I don't mind acting as one of those idiots. I figure if most of my favourite bands can't read music, there's interesting math to be done by people who aren't proving anything or making progress, but intentionally or not, like all idiots, they exist as an example to inspire others. :)