java - Artificial Neural Networks input comprehension -
i'm total beginner regarding a.n.n.s. understand concept , there's no straight explanation why input series of 0s , 1s , output series of 0s , 1s.
i read here on neural networks - input values can encode input data normalization function it's converted number between 0 , 1.
is case or misunderstanding things?
also think point me in right direction regarding article/ lecturing material should pick clear things out?
i'm relearning nets now, , asked similar question.
it's hard know exact scenario is, me, activations in range of 0-1 because activation function sigmoid function, outputs in range of 0-1 (although you'll need ask math oriented person why is).
say you're using simple step function instead activation function. logically take 0 or 1 (but nothing in between), , output 0 or 1.
so answer seems be: range of activations in net defined activation function used.
Comments
Post a Comment