First time here? Checkout the FAQ!
x
+3 votes
4k views
asked in Machine Learning by (116k points)  

The goal of backpropagation is to optimize the weights so that the neural network can learn how to correctly map arbitrary inputs to outputs.

Assume for the following neural network, inputs = [$i_1,i_2$] = [0.05, 0.10], we want the neural network to output = [$o_1$,$o_2$] = [0.01, 0.99], and for learning rate, $\alpha=0.5$.
In addition, the activation function for the hidden layer (both $h_1$ and $h_2$) is sigmoid (logistic):

$S(x)=\frac{1}{1+e^{-x}}$

Hint:
$w_{new} = w_{old} - \alpha \frac{\partial E}{\partial w}$

$E_{\text {total}}=\sum \frac{1}{2}(\text {target}-\text {output})^{2}$

 

a) Show step by step solution to calculate weights $w_1$ to $w_8$ after one update in table below.
b) Calculate initial error and error after one update (assume biases $[b_1,b_2]$ are not changing during the updates).

Updating weights in backpropagation algorithm
Weights Initialization New weights after one step
$w1$ 0.15 ?
$w2$ 0.20 ?
$w3$ 0.25 ?
$w4$ 0.30 ?
$w5$ 0.40 ?
$w6$ 0.45 ?
$w7$ 0.50 ?
$w8$ 0.55 ?
  

1 Answer

+2 votes
answered by (116k points)  
selected by
 
Best answer

You can find the step by step solution which is provided here.

commented by (100 points)  
While the question says activation function for hidden layer, the solution applies the same activation function to output layer as well. Do we also need to apply activation function to output layer?
...