The correspondence relationship is as follows:
Code inside dprice 1 Why is it?
Because we know the actual neural networks, back-propagation of the first item is
∂y2∂E
So here for the convenience of the reader's understanding, there is no concept of E leads.
So the meaning here is dprice
Y2Y2
So the above are:
x2Y2=1.1
dapple_price=∂y2∂y2⋅∂x2∂y2=1⋅1.1=1.1
dapple=∂y2∂y2⋅∂x2∂y2⋅∂y1∂x2⋅∂x1∂y1=1⋅1.1⋅1⋅2=2.2
Note that there is still not related to the introduction of amendments to calculate weights, so in fact and real neural networks were far worse