3 ms·
If you are interested, I would suggest reading up on the https://en.wikipedia.org/wiki/Universal_approximation_theorem https://en.wikipedia.org/wiki/Universal_a
by sjg007 7y ago
If you are interested, I would suggest reading up on the https://en.wikipedia.org/wiki/Universal_approximation_theorem https://en.wikipedia.org/wiki/Universal_approximation_theore...
- 6gvONxR4sf7o 7y agoThere are theorems like that for polynomials and fourier series and all sorts of other function classes too. They are just as practically relevant (or irrelevant).
- sjg007 7y agoSure... I mean it is matrices all the way down but the claim that AI (e.g. deep learning) is just applied statistics is disingenuous.