So a "neural network" is actually a type of parameterized mathematical function ... | Hacker News

HN2new | past | comments | ask | show | jobs | submit

		api on Dec 6, 2020 \| parent \| context \| favorite \| on: Every Model Learned by Gradient Descent Is Approxi... So a "neural network" is actually a type of parameterized mathematical function that can be fit to any curve including higher dimensional surfaces, etc.?

xksteven on Dec 6, 2020 | [–]

That's correct. See https://en.wikipedia.org/wiki/Universal_approximation_theore... for more details

laingc on Dec 6, 2020 | [–]

Yes.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact