Skip to content

Comment on Ask HN: What happened to fuzzy logic?parent

Comments

There was a popular conception (like taught in textbooks even) that one hidden layer was enough since a neural network with one hidden layer could approximate any function. Deep learning became a thing when it was found to work great in practice.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.