Why you multiply the sigmoid_prime(x) by x? You defined backpropagation error as (𝑦−𝑦̂)𝜎′(𝑥) in the notebook:
Your solution:
def error_term_formula(x, y, output):
# for binary cross entropy loss
return (y - output)*x
# for mean square error
# return (y - output)*sigmoid_prime(x)*x
Notebook definition:
TODO: Backpropagate the error
Now it's your turn to shine. Write the error term. Remember that this is given by the equation
(𝑦−𝑦̂)𝜎′(𝑥)
You even had (y - output)*sigmoid_prime(x) as the solution, but then you changed it to (y - output)*sigmoid_prime(x)*x in this commit
Why you multiply the sigmoid_prime(x) by x? You defined backpropagation error as
(𝑦−𝑦̂)𝜎′(𝑥)
in the notebook: Your solution:Notebook definition: