Output for an RNN

问题

I'm taking the Coursera course Neural Networks for Machine Learning hosted by Geoffrey Hinton from the University of Toronto and there is a quiz question in week 7 for which my answer differs from the right one.

The question goes like this:

One question is, how should I get a probability between 0 and 1 if the Whh weight is negative and the logistic h unit gives values between 0 and 1. Given the above, their linear combination will allways be negative.

A second question would be if we also have to use backpropagation in order to get the right answer?

The way I've started to tackle this question is the following:

h0 = 1/( 1 + exp(- (Whh * hbias + Wxh*x0)) )

h1 = 1/( 1 + exp(- (Whh * h0 + Wxh*x1)))

y1 = Why * h1

Which of my assumptions are incorrect?

Regards, Cristian

回答1:

Try this.

h0 = sigmoid(x0*Wxh + hbias)
h1 = sigmoid(x1*Wxh + Whh*h0 + hbias)
y1 = h1*Why

来源：https://stackoverflow.com/questions/43044992/output-for-an-rnn