I have done this a lot, focus on mathematical concepts more, less on implementation. I can confirm that this is not the right approach for industry, may be fine for research
I’m not going to speak for the author here but from my experience, I would say pretty much all ML textbooks are outdated at this point in time. I personally think the best way to learn ML, especially in the applied context, is by going through Andrej Karpathy’s YouTube series. Watching him build LLMs, neural networks, etc from scratch is an incredibly helpful exercise for understanding deep learning.
I have done this a lot, focus on mathematical concepts more, less on implementation. I can confirm that this is not the right approach for industry, may be fine for research
My suggestion watch a course by any prof and solve the exercises. Or pick any book but complete the exercises
Is there a particular book on the subject you would recommend? Thank you.
I’m not going to speak for the author here but from my experience, I would say pretty much all ML textbooks are outdated at this point in time. I personally think the best way to learn ML, especially in the applied context, is by going through Andrej Karpathy’s YouTube series. Watching him build LLMs, neural networks, etc from scratch is an incredibly helpful exercise for understanding deep learning.