Backpropagation Algorithm
From Ufldl
m |
|||
Line 85: | Line 85: | ||
compute all the activations throughout the network, including the output value | compute all the activations throughout the network, including the output value | ||
of the hypothesis <math>h_{W,b}(x)</math>. Then, for each node <math>i</math> in layer <math>l</math>, we would like | of the hypothesis <math>h_{W,b}(x)</math>. Then, for each node <math>i</math> in layer <math>l</math>, we would like | ||
- | to compute an | + | to compute an "error term" <math>\delta^{(l)}_i</math> that measures how much that node was |
- | + | "responsible" for any errors in our output. | |
For an output node, we can directly measure the difference between the | For an output node, we can directly measure the difference between the | ||
network's activation and the true target value, and use that to define | network's activation and the true target value, and use that to define |