1. Linear Algebra (The Language of Data)

   When working with machine learning models or neural networks, data is almost always represented as matrices and vectors. You don’t need to write complex proofs, but you should understand how data is manipulated.

2. Calculus (The Engine of Learning)

   You won’t necessarily be calculating complex integrals by hand, but you need to understand the language of change, because optimizing an AI model (like training a neural network via backpropagation) relies entirely on calculus.

3. Probability and Core Statistics (The Foundation of Analytics)

For (1) Khan Academy is an excellent source for targeted brush-ups on Linear Algebra and Multivariable Calculus. Also, MIT OpenCourseWare (18.06 Linear Algebra by Gilbert Strang) is free to download.

If you are a visual learner then StatQuest with Josh Starmer (YouTube): Great for visual, intuitive explanations of core statistics and machine learning concepts without overwhelming mathematical jargon.

For (2) and (3) students have stated that 3Blue1Brown (YouTube) has playlists “Essence of Linear Algebra” and “Essence of Calculus”. They provide the geometric, intuitive understanding that makes the coding side of AI much easier to grasp. Also, OpenIntro Statistics (David Diez et al.).