This is a question almost everyone has when starting out or transitioning into Machine Learning. I know there are many answers and different perspectives (I’ve seen plenty of YouTube suggestions). But I’d like to make this question a bit more general.

For someone who is interested in or wants to build a career in Machine Learning, what level of math should they learn at the beginner stage? For example, while learning basic Machine Learning, Boosting Techniques, Feature Scaling, and more, what topics should they focus on to build a strong foundation?

Feel free to include what you consider the basics of Machine Learning. All suggestions are welcome!

I’d say you need this much math if you want to be a practitioner:

You’d need more if you’re aiming to get into research.

For context, I didn’t learn linear algebra (eigenvalues, eigenvectors, Wronskian, gradient descent, etc.) until after I completed Calc 1 (AP Calc AB), Calc 2 (AP Calc BC), and Calc 3 at university.

@Rayne
I’m aiming to be an ML Engineer or Data Scientist. I didn’t quite get what you meant by ‘practitioner,’ though. Your resource (the Reddit post) is helpful for answering the question. Thanks!

Darwin said: @Rayne
I’m aiming to be an ML Engineer or Data Scientist. I didn’t quite get what you meant by ‘practitioner,’ though. Your resource (the Reddit post) is helpful for answering the question. Thanks!

That’s what I meant by ‘practitioner’—someone who works hands-on with ML. There are also online certificates available at:

Just keep in mind, in practice, employers often look for a degree and some experience. It can be tough to land a job if you’re completely self-taught.