21–30 of 95 posts
Paradigms of Artificial Intelligence Programming (PAIP) One of the best books on AI and Programming ever.
Paradigms of Artificial Intelligence Programming (PAIP) One of the best books on AI and Programming ever.
Unfortunately, there is not a drop of numerics in that book. It's a good book for learning about symbolic AI. There is, to a solid first order approximation, zero symbolic AI in a system like, say, Google speech recognition.
To iterate on what others said, but what was not emphasized enough from my point of few: AI is academic (as a synonym for 'theoretical' and 'math-intensive'). Once you look beyond purely symbolic AI, which proved to be infeasible as @curuinor pointed out somewhere here, you will need to build up at least basic knowledge in probability theory and linear algebra. The path I'm following at the moment is a quite rigorous…
1. AI: A Modern Approach by Stuart Russell and Peter Norvig.
2. Deep Learning by Ian Goodfellow and Yoshua Bengio.
It is amazing how approachable both books are for beginners, but you will be diving a lot into academic stuff as you go along.
The challenge I've found with books is the space has been moving so quickly in the past 10 years. By the time the book is out, the methods described in it are no longer state of the art.
To iterate on what others said, but what was not emphasized enough from my point of few: AI is academic (as a synonym for 'theoretical' and 'math-intensive'). Once you look beyond purely symbolic AI, which proved to be infeasible as @curuinor pointed out somewhere here, you will need to build up at least basic knowledge in probability theory and linear algebra. The path I'm following at the moment is a quite rigorous…
Now, from the link: "Few universities offer an education that is on par with what you can find online these days. The people pioneering the field from industry and academia so openly and competently share their knowledge that the best curriculum is an open source one."
On the one hand, it is true there are a ton of resources where the largest cost is the time it takes to go through the learning process. And I'm awestruck that research papers are so openly available and practitioners are so willing to share their knowledge to others both in posting their books as PDFs/HTML files and creating online courses.
On the other hand, how feasible is it for an individual to work on notable AI companies/projects without a Masters or PhD in a related field? Can that gap be crossed merely by becoming fluent in the various disciplines involved in AI, before contributing non-formally academic research/experiments you've conducted on your own?
- The Master Algorithm: made for a general audience, gives you a lay of the land
- Python Machine Learning by Sebastian Raschka: gives you practical skills using python, scikit-learn, numpy, jupyter notebooks, pandas etc. From zero to kaggle in 4 chapters, goes deeper after that. Also goes into enough theory you aren't flying completely blind.
After that, I'm afraid I think you do need to go "academic", if by that you mean learning some of the underlying math to approach AI / ML from a more rigorous probabilistic perspective. I'd recommend studying probability theory and then working your way through Bishop's Pattern Recognition and Machine Learning. After that a lot more doors open up too more specialized topics like computer vision, reinforcement learning etc.
I've written up a lot more about this here: