When I was first getting into Deep Learning, learning the proof of the universal approximation theorem helped a lot. Once you understand why neural networks ar

When I was first getting into Deep Learning, learning the proof of the universal approximation theorem helped a lot. Once you understand why neural networks are able to approximate functions, it makes everything built on top of them much easier to understand.