It might feel like Machine Learning (ML) has been around forever, but ML with Neural Networks (NNs) only became mainstream in the 2010s. A great example is the well-known AlexNet, developed by Alex Krizhevsky, Ilya Sutskever, and Geoffrey Hinton, which won the ImageNet competition by a significant margin in 2012. That was a little more than ten years ago.
Back in college, from 2007 to 2009, I studied linear algebra, calculus, probability, and statistics. At that time, it wasn’t common to talk about applying these subjects to Neural Networks or even run computational experiments. We approached everything analytically using pen and paper, with only a few minor exceptions.
One of the challenges I faced while learning about ML and NNs was the terminology. I discovered that many math terms, theorem names, notations, and even certain semantic structures don’t always translate well between languages. This was quite different from Computer Science, where many terms in my native language are borrowed directly from English, making it easier to follow along.
I always try to grasp the foundational concepts of mathematics for ML at least once. It’s okay if I forget some of it later. If I can understand it well the first time, I know I’ll be able to reconstruct my understanding using resources like the Internet and books whenever I need to. Take backpropagation, for example. While I don’t see much practical value in being able to implement it from scratch (unless you’re just graduating and looking for a junior ML position), I think it’s important to do it at least once, even if you’re using advanced libraries like NumPy. It’s all about building a deeper understanding.
So I decided to feed a few birds with one scone: re-learn all the necessary math in English to connect the terminology, understand every component at least once, and link abstract techniques and tools directly to ML rather than keeping them abstract. I chose the “Mathematics for Machine Learning and Data Science” specialization from DeepLearning.AI, and finally - I finished it! 🎉