Ask HN: The ML/AI space is overwhelmingly large. Where to start?

[triyambakam]

I'm a "traditional" software engineer with many years of experience in web technologies. I feel fascinated by ML/AI technologies, but I don't know where to begin to learn about it. I am also not even sure yet if it's something I would enjoy, so I want to find the best way to maximize my investment of time vs return in experience and skill. I am specifically interested in computer vision and natural language processing. Where should I begin? Is Andrew Ng's course still a good starting point? Or the D2L.ai course?

[ak_111]

It's almost impossible having a decent understanding of the theoretical side of ML without having a decent understanding of linear algebra, multivariable calculus and probability. So you need to start with those if you haven't the background and most ML courses (including Ng's) assume you know some of that. Going through the undergraduate material for these three topics is probably sufficient.