Suppose that
[0.4
0.3 0.9 0.3b"=1,1,1); and
0.2
2
= 0.2, b²=0.5).
0.8
If the input is a =1). what is the network output? Show your calculation steps and round your
answer to 4 decimals.
[Answer]
Q1.2: Backward Propagation
Use the chain rule to derive the expressions of the following gradients:
1.
2.
ƏL
dwz
ƏL
owl
and
and
ƏL
ab²)
ƏL
ab¹
Your final answers should only include the variables appeared in the question.
Hint #1: Begin by writing down the chain of partial derivatives, and then plug in predefined
variables.
Hint #2: While plugging in predefined variables, be careful about the dimensions and
orientation. You can first write down the expressions in the element level and then figure out the
matrix form.
Hint #3: The derivative of a (x) is a (x)(1-o(x)).
Fig: 1